* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
- * or http://www.opensolaris.org/os/licensing.
+ * or https://opensource.org/licenses/CDDL-1.0.
* See the License for the specific language governing permissions
* and limitations under the License.
*
*/
/*
* Copyright (c) 2005, 2010, Oracle and/or its affiliates. All rights reserved.
- * Copyright (c) 2012, 2015 by Delphix. All rights reserved.
+ * Copyright (c) 2012, 2020 by Delphix. All rights reserved.
* Copyright (c) 2014 Spectra Logic Corporation, All rights reserved.
*/
#include <sys/zio.h>
#include <sys/dmu_zfetch.h>
#include <sys/range_tree.h>
-#include <sys/trace_dnode.h>
-
+#include <sys/trace_zfs.h>
+#include <sys/zfs_project.h>
+
+dnode_stats_t dnode_stats = {
+ { "dnode_hold_dbuf_hold", KSTAT_DATA_UINT64 },
+ { "dnode_hold_dbuf_read", KSTAT_DATA_UINT64 },
+ { "dnode_hold_alloc_hits", KSTAT_DATA_UINT64 },
+ { "dnode_hold_alloc_misses", KSTAT_DATA_UINT64 },
+ { "dnode_hold_alloc_interior", KSTAT_DATA_UINT64 },
+ { "dnode_hold_alloc_lock_retry", KSTAT_DATA_UINT64 },
+ { "dnode_hold_alloc_lock_misses", KSTAT_DATA_UINT64 },
+ { "dnode_hold_alloc_type_none", KSTAT_DATA_UINT64 },
+ { "dnode_hold_free_hits", KSTAT_DATA_UINT64 },
+ { "dnode_hold_free_misses", KSTAT_DATA_UINT64 },
+ { "dnode_hold_free_lock_misses", KSTAT_DATA_UINT64 },
+ { "dnode_hold_free_lock_retry", KSTAT_DATA_UINT64 },
+ { "dnode_hold_free_overflow", KSTAT_DATA_UINT64 },
+ { "dnode_hold_free_refcount", KSTAT_DATA_UINT64 },
+ { "dnode_free_interior_lock_retry", KSTAT_DATA_UINT64 },
+ { "dnode_allocate", KSTAT_DATA_UINT64 },
+ { "dnode_reallocate", KSTAT_DATA_UINT64 },
+ { "dnode_buf_evict", KSTAT_DATA_UINT64 },
+ { "dnode_alloc_next_chunk", KSTAT_DATA_UINT64 },
+ { "dnode_alloc_race", KSTAT_DATA_UINT64 },
+ { "dnode_alloc_next_block", KSTAT_DATA_UINT64 },
+ { "dnode_move_invalid", KSTAT_DATA_UINT64 },
+ { "dnode_move_recheck1", KSTAT_DATA_UINT64 },
+ { "dnode_move_recheck2", KSTAT_DATA_UINT64 },
+ { "dnode_move_special", KSTAT_DATA_UINT64 },
+ { "dnode_move_handle", KSTAT_DATA_UINT64 },
+ { "dnode_move_rwlock", KSTAT_DATA_UINT64 },
+ { "dnode_move_active", KSTAT_DATA_UINT64 },
+};
+
+dnode_sums_t dnode_sums;
+
+static kstat_t *dnode_ksp;
static kmem_cache_t *dnode_cache;
-/*
- * Define DNODE_STATS to turn on statistic gathering. By default, it is only
- * turned on when DEBUG is also defined.
- */
-#ifdef DEBUG
-#define DNODE_STATS
-#endif /* DEBUG */
-
-#ifdef DNODE_STATS
-#define DNODE_STAT_ADD(stat) ((stat)++)
-#else
-#define DNODE_STAT_ADD(stat) /* nothing */
-#endif /* DNODE_STATS */
-ASSERTV(static dnode_phys_t dnode_phys_zero);
+static dnode_phys_t dnode_phys_zero __maybe_unused;
int zfs_default_bs = SPA_MINBLOCKSHIFT;
int zfs_default_ibs = DN_MAX_INDBLKSHIFT;
const dmu_buf_impl_t *d1 = x1;
const dmu_buf_impl_t *d2 = x2;
- if (d1->db_level < d2->db_level) {
- return (-1);
- }
- if (d1->db_level > d2->db_level) {
- return (1);
- }
+ int cmp = TREE_CMP(d1->db_level, d2->db_level);
+ if (likely(cmp))
+ return (cmp);
- if (d1->db_blkid < d2->db_blkid) {
- return (-1);
- }
- if (d1->db_blkid > d2->db_blkid) {
- return (1);
+ cmp = TREE_CMP(d1->db_blkid, d2->db_blkid);
+ if (likely(cmp))
+ return (cmp);
+
+ if (d1->db_state == DB_MARKER) {
+ ASSERT3S(d2->db_state, !=, DB_MARKER);
+ return (TREE_PCMP(d1->db_parent, d2));
+ } else if (d2->db_state == DB_MARKER) {
+ ASSERT3S(d1->db_state, !=, DB_MARKER);
+ return (TREE_PCMP(d1, d2->db_parent));
}
if (d1->db_state == DB_SEARCH) {
return (1);
}
- if ((uintptr_t)d1 < (uintptr_t)d2) {
- return (-1);
- }
- if ((uintptr_t)d1 > (uintptr_t)d2) {
- return (1);
- }
- return (0);
+ return (TREE_PCMP(d1, d2));
}
-/* ARGSUSED */
static int
dnode_cons(void *arg, void *unused, int kmflag)
{
+ (void) unused, (void) kmflag;
dnode_t *dn = arg;
- int i;
- rw_init(&dn->dn_struct_rwlock, NULL, RW_DEFAULT, NULL);
+ rw_init(&dn->dn_struct_rwlock, NULL, RW_NOLOCKDEP, NULL);
mutex_init(&dn->dn_mtx, NULL, MUTEX_DEFAULT, NULL);
mutex_init(&dn->dn_dbufs_mtx, NULL, MUTEX_DEFAULT, NULL);
cv_init(&dn->dn_notxholds, NULL, CV_DEFAULT, NULL);
+ cv_init(&dn->dn_nodnholds, NULL, CV_DEFAULT, NULL);
/*
* Every dbuf has a reference, and dropping a tracked reference is
* O(number of references), so don't track dn_holds.
*/
- refcount_create_untracked(&dn->dn_holds);
- refcount_create(&dn->dn_tx_holds);
+ zfs_refcount_create_untracked(&dn->dn_holds);
+ zfs_refcount_create(&dn->dn_tx_holds);
list_link_init(&dn->dn_link);
- bzero(&dn->dn_next_nblkptr[0], sizeof (dn->dn_next_nblkptr));
- bzero(&dn->dn_next_nlevels[0], sizeof (dn->dn_next_nlevels));
- bzero(&dn->dn_next_indblkshift[0], sizeof (dn->dn_next_indblkshift));
- bzero(&dn->dn_next_bonustype[0], sizeof (dn->dn_next_bonustype));
- bzero(&dn->dn_rm_spillblk[0], sizeof (dn->dn_rm_spillblk));
- bzero(&dn->dn_next_bonuslen[0], sizeof (dn->dn_next_bonuslen));
- bzero(&dn->dn_next_blksz[0], sizeof (dn->dn_next_blksz));
-
- for (i = 0; i < TXG_SIZE; i++) {
- list_link_init(&dn->dn_dirty_link[i]);
+ memset(dn->dn_next_type, 0, sizeof (dn->dn_next_type));
+ memset(dn->dn_next_nblkptr, 0, sizeof (dn->dn_next_nblkptr));
+ memset(dn->dn_next_nlevels, 0, sizeof (dn->dn_next_nlevels));
+ memset(dn->dn_next_indblkshift, 0, sizeof (dn->dn_next_indblkshift));
+ memset(dn->dn_next_bonustype, 0, sizeof (dn->dn_next_bonustype));
+ memset(dn->dn_rm_spillblk, 0, sizeof (dn->dn_rm_spillblk));
+ memset(dn->dn_next_bonuslen, 0, sizeof (dn->dn_next_bonuslen));
+ memset(dn->dn_next_blksz, 0, sizeof (dn->dn_next_blksz));
+ memset(dn->dn_next_maxblkid, 0, sizeof (dn->dn_next_maxblkid));
+
+ for (int i = 0; i < TXG_SIZE; i++) {
+ multilist_link_init(&dn->dn_dirty_link[i]);
dn->dn_free_ranges[i] = NULL;
list_create(&dn->dn_dirty_records[i],
sizeof (dbuf_dirty_record_t),
dn->dn_allocated_txg = 0;
dn->dn_free_txg = 0;
dn->dn_assigned_txg = 0;
+ dn->dn_dirty_txg = 0;
dn->dn_dirtyctx = 0;
dn->dn_dirtyctx_firstset = NULL;
dn->dn_bonus = NULL;
dn->dn_oldflags = 0;
dn->dn_olduid = 0;
dn->dn_oldgid = 0;
+ dn->dn_oldprojid = ZFS_DEFAULT_PROJID;
dn->dn_newuid = 0;
dn->dn_newgid = 0;
+ dn->dn_newprojid = ZFS_DEFAULT_PROJID;
dn->dn_id_flags = 0;
dn->dn_dbufs_count = 0;
- dn->dn_unlisted_l0_blkid = 0;
avl_create(&dn->dn_dbufs, dbuf_compare, sizeof (dmu_buf_impl_t),
offsetof(dmu_buf_impl_t, db_link));
return (0);
}
-/* ARGSUSED */
static void
dnode_dest(void *arg, void *unused)
{
- int i;
+ (void) unused;
dnode_t *dn = arg;
rw_destroy(&dn->dn_struct_rwlock);
mutex_destroy(&dn->dn_mtx);
mutex_destroy(&dn->dn_dbufs_mtx);
cv_destroy(&dn->dn_notxholds);
- refcount_destroy(&dn->dn_holds);
- refcount_destroy(&dn->dn_tx_holds);
+ cv_destroy(&dn->dn_nodnholds);
+ zfs_refcount_destroy(&dn->dn_holds);
+ zfs_refcount_destroy(&dn->dn_tx_holds);
ASSERT(!list_link_active(&dn->dn_link));
- for (i = 0; i < TXG_SIZE; i++) {
- ASSERT(!list_link_active(&dn->dn_dirty_link[i]));
+ for (int i = 0; i < TXG_SIZE; i++) {
+ ASSERT(!multilist_link_active(&dn->dn_dirty_link[i]));
ASSERT3P(dn->dn_free_ranges[i], ==, NULL);
list_destroy(&dn->dn_dirty_records[i]);
ASSERT0(dn->dn_next_nblkptr[i]);
ASSERT0(dn->dn_rm_spillblk[i]);
ASSERT0(dn->dn_next_bonuslen[i]);
ASSERT0(dn->dn_next_blksz[i]);
+ ASSERT0(dn->dn_next_maxblkid[i]);
}
ASSERT0(dn->dn_allocated_txg);
ASSERT0(dn->dn_free_txg);
ASSERT0(dn->dn_assigned_txg);
+ ASSERT0(dn->dn_dirty_txg);
ASSERT0(dn->dn_dirtyctx);
ASSERT3P(dn->dn_dirtyctx_firstset, ==, NULL);
ASSERT3P(dn->dn_bonus, ==, NULL);
ASSERT0(dn->dn_oldflags);
ASSERT0(dn->dn_olduid);
ASSERT0(dn->dn_oldgid);
+ ASSERT0(dn->dn_oldprojid);
ASSERT0(dn->dn_newuid);
ASSERT0(dn->dn_newgid);
+ ASSERT0(dn->dn_newprojid);
ASSERT0(dn->dn_id_flags);
ASSERT0(dn->dn_dbufs_count);
- ASSERT0(dn->dn_unlisted_l0_blkid);
avl_destroy(&dn->dn_dbufs);
}
+static int
+dnode_kstats_update(kstat_t *ksp, int rw)
+{
+ dnode_stats_t *ds = ksp->ks_data;
+
+ if (rw == KSTAT_WRITE)
+ return (EACCES);
+ ds->dnode_hold_dbuf_hold.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_hold_dbuf_hold);
+ ds->dnode_hold_dbuf_read.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_hold_dbuf_read);
+ ds->dnode_hold_alloc_hits.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_hold_alloc_hits);
+ ds->dnode_hold_alloc_misses.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_hold_alloc_misses);
+ ds->dnode_hold_alloc_interior.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_hold_alloc_interior);
+ ds->dnode_hold_alloc_lock_retry.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_hold_alloc_lock_retry);
+ ds->dnode_hold_alloc_lock_misses.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_hold_alloc_lock_misses);
+ ds->dnode_hold_alloc_type_none.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_hold_alloc_type_none);
+ ds->dnode_hold_free_hits.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_hold_free_hits);
+ ds->dnode_hold_free_misses.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_hold_free_misses);
+ ds->dnode_hold_free_lock_misses.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_hold_free_lock_misses);
+ ds->dnode_hold_free_lock_retry.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_hold_free_lock_retry);
+ ds->dnode_hold_free_refcount.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_hold_free_refcount);
+ ds->dnode_hold_free_overflow.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_hold_free_overflow);
+ ds->dnode_free_interior_lock_retry.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_free_interior_lock_retry);
+ ds->dnode_allocate.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_allocate);
+ ds->dnode_reallocate.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_reallocate);
+ ds->dnode_buf_evict.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_buf_evict);
+ ds->dnode_alloc_next_chunk.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_alloc_next_chunk);
+ ds->dnode_alloc_race.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_alloc_race);
+ ds->dnode_alloc_next_block.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_alloc_next_block);
+ ds->dnode_move_invalid.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_move_invalid);
+ ds->dnode_move_recheck1.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_move_recheck1);
+ ds->dnode_move_recheck2.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_move_recheck2);
+ ds->dnode_move_special.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_move_special);
+ ds->dnode_move_handle.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_move_handle);
+ ds->dnode_move_rwlock.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_move_rwlock);
+ ds->dnode_move_active.value.ui64 =
+ wmsum_value(&dnode_sums.dnode_move_active);
+ return (0);
+}
+
void
dnode_init(void)
{
dnode_cache = kmem_cache_create("dnode_t", sizeof (dnode_t),
0, dnode_cons, dnode_dest, NULL, NULL, NULL, 0);
kmem_cache_set_move(dnode_cache, dnode_move);
+
+ wmsum_init(&dnode_sums.dnode_hold_dbuf_hold, 0);
+ wmsum_init(&dnode_sums.dnode_hold_dbuf_read, 0);
+ wmsum_init(&dnode_sums.dnode_hold_alloc_hits, 0);
+ wmsum_init(&dnode_sums.dnode_hold_alloc_misses, 0);
+ wmsum_init(&dnode_sums.dnode_hold_alloc_interior, 0);
+ wmsum_init(&dnode_sums.dnode_hold_alloc_lock_retry, 0);
+ wmsum_init(&dnode_sums.dnode_hold_alloc_lock_misses, 0);
+ wmsum_init(&dnode_sums.dnode_hold_alloc_type_none, 0);
+ wmsum_init(&dnode_sums.dnode_hold_free_hits, 0);
+ wmsum_init(&dnode_sums.dnode_hold_free_misses, 0);
+ wmsum_init(&dnode_sums.dnode_hold_free_lock_misses, 0);
+ wmsum_init(&dnode_sums.dnode_hold_free_lock_retry, 0);
+ wmsum_init(&dnode_sums.dnode_hold_free_refcount, 0);
+ wmsum_init(&dnode_sums.dnode_hold_free_overflow, 0);
+ wmsum_init(&dnode_sums.dnode_free_interior_lock_retry, 0);
+ wmsum_init(&dnode_sums.dnode_allocate, 0);
+ wmsum_init(&dnode_sums.dnode_reallocate, 0);
+ wmsum_init(&dnode_sums.dnode_buf_evict, 0);
+ wmsum_init(&dnode_sums.dnode_alloc_next_chunk, 0);
+ wmsum_init(&dnode_sums.dnode_alloc_race, 0);
+ wmsum_init(&dnode_sums.dnode_alloc_next_block, 0);
+ wmsum_init(&dnode_sums.dnode_move_invalid, 0);
+ wmsum_init(&dnode_sums.dnode_move_recheck1, 0);
+ wmsum_init(&dnode_sums.dnode_move_recheck2, 0);
+ wmsum_init(&dnode_sums.dnode_move_special, 0);
+ wmsum_init(&dnode_sums.dnode_move_handle, 0);
+ wmsum_init(&dnode_sums.dnode_move_rwlock, 0);
+ wmsum_init(&dnode_sums.dnode_move_active, 0);
+
+ dnode_ksp = kstat_create("zfs", 0, "dnodestats", "misc",
+ KSTAT_TYPE_NAMED, sizeof (dnode_stats) / sizeof (kstat_named_t),
+ KSTAT_FLAG_VIRTUAL);
+ if (dnode_ksp != NULL) {
+ dnode_ksp->ks_data = &dnode_stats;
+ dnode_ksp->ks_update = dnode_kstats_update;
+ kstat_install(dnode_ksp);
+ }
}
void
dnode_fini(void)
{
+ if (dnode_ksp != NULL) {
+ kstat_delete(dnode_ksp);
+ dnode_ksp = NULL;
+ }
+
+ wmsum_fini(&dnode_sums.dnode_hold_dbuf_hold);
+ wmsum_fini(&dnode_sums.dnode_hold_dbuf_read);
+ wmsum_fini(&dnode_sums.dnode_hold_alloc_hits);
+ wmsum_fini(&dnode_sums.dnode_hold_alloc_misses);
+ wmsum_fini(&dnode_sums.dnode_hold_alloc_interior);
+ wmsum_fini(&dnode_sums.dnode_hold_alloc_lock_retry);
+ wmsum_fini(&dnode_sums.dnode_hold_alloc_lock_misses);
+ wmsum_fini(&dnode_sums.dnode_hold_alloc_type_none);
+ wmsum_fini(&dnode_sums.dnode_hold_free_hits);
+ wmsum_fini(&dnode_sums.dnode_hold_free_misses);
+ wmsum_fini(&dnode_sums.dnode_hold_free_lock_misses);
+ wmsum_fini(&dnode_sums.dnode_hold_free_lock_retry);
+ wmsum_fini(&dnode_sums.dnode_hold_free_refcount);
+ wmsum_fini(&dnode_sums.dnode_hold_free_overflow);
+ wmsum_fini(&dnode_sums.dnode_free_interior_lock_retry);
+ wmsum_fini(&dnode_sums.dnode_allocate);
+ wmsum_fini(&dnode_sums.dnode_reallocate);
+ wmsum_fini(&dnode_sums.dnode_buf_evict);
+ wmsum_fini(&dnode_sums.dnode_alloc_next_chunk);
+ wmsum_fini(&dnode_sums.dnode_alloc_race);
+ wmsum_fini(&dnode_sums.dnode_alloc_next_block);
+ wmsum_fini(&dnode_sums.dnode_move_invalid);
+ wmsum_fini(&dnode_sums.dnode_move_recheck1);
+ wmsum_fini(&dnode_sums.dnode_move_recheck2);
+ wmsum_fini(&dnode_sums.dnode_move_special);
+ wmsum_fini(&dnode_sums.dnode_move_handle);
+ wmsum_fini(&dnode_sums.dnode_move_rwlock);
+ wmsum_fini(&dnode_sums.dnode_move_active);
+
kmem_cache_destroy(dnode_cache);
dnode_cache = NULL;
}
}
if (dn->dn_phys->dn_type != DMU_OT_NONE || dn->dn_allocated_txg != 0) {
int i;
+ int max_bonuslen = DN_SLOTS_TO_BONUSLEN(dn->dn_num_slots);
ASSERT3U(dn->dn_indblkshift, <=, SPA_MAXBLOCKSHIFT);
if (dn->dn_datablkshift) {
ASSERT3U(dn->dn_datablkshift, >=, SPA_MINBLOCKSHIFT);
ASSERT(DMU_OT_IS_VALID(dn->dn_type));
ASSERT3U(dn->dn_nblkptr, >=, 1);
ASSERT3U(dn->dn_nblkptr, <=, DN_MAX_NBLKPTR);
- ASSERT3U(dn->dn_bonuslen, <=, DN_MAX_BONUSLEN);
+ ASSERT3U(dn->dn_bonuslen, <=, max_bonuslen);
ASSERT3U(dn->dn_datablksz, ==,
dn->dn_datablkszsec << SPA_MINBLOCKSHIFT);
ASSERT3U(ISP2(dn->dn_datablksz), ==, dn->dn_datablkshift != 0);
ASSERT3U((dn->dn_nblkptr - 1) * sizeof (blkptr_t) +
- dn->dn_bonuslen, <=, DN_MAX_BONUSLEN);
+ dn->dn_bonuslen, <=, max_bonuslen);
for (i = 0; i < TXG_SIZE; i++) {
ASSERT3U(dn->dn_next_nlevels[i], <=, dn->dn_nlevels);
}
int i;
if (dnp->dn_type == DMU_OT_NONE) {
- bzero(dnp, sizeof (dnode_phys_t));
+ memset(dnp, 0, sizeof (dnode_phys_t));
return;
}
dnp->dn_datablkszsec = BSWAP_16(dnp->dn_datablkszsec);
dnp->dn_bonuslen = BSWAP_16(dnp->dn_bonuslen);
+ dnp->dn_extra_slots = BSWAP_8(dnp->dn_extra_slots);
dnp->dn_maxblkid = BSWAP_64(dnp->dn_maxblkid);
dnp->dn_used = BSWAP_64(dnp->dn_used);
* dnode dnode is smaller than a regular dnode.
*/
if (dnp->dn_bonuslen != 0) {
- /*
- * Note that the bonus length calculated here may be
- * longer than the actual bonus buffer. This is because
- * we always put the bonus buffer after the last block
- * pointer (instead of packing it against the end of the
- * dnode buffer).
- */
- int off = (dnp->dn_nblkptr-1) * sizeof (blkptr_t);
- size_t len = DN_MAX_BONUSLEN - off;
dmu_object_byteswap_t byteswap;
ASSERT(DMU_OT_IS_VALID(dnp->dn_bonustype));
byteswap = DMU_OT_BYTESWAP(dnp->dn_bonustype);
- dmu_ot_byteswap[byteswap].ob_func(dnp->dn_bonus + off, len);
+ dmu_ot_byteswap[byteswap].ob_func(DN_BONUS(dnp),
+ DN_MAX_BONUS_LEN(dnp));
}
/* Swap SPILL block if we have one */
if (dnp->dn_flags & DNODE_FLAG_SPILL_BLKPTR)
- byteswap_uint64_array(&dnp->dn_spill, sizeof (blkptr_t));
-
+ byteswap_uint64_array(DN_SPILL_BLKPTR(dnp), sizeof (blkptr_t));
}
void
dnode_buf_byteswap(void *vbuf, size_t size)
{
- dnode_phys_t *buf = vbuf;
- int i;
+ int i = 0;
ASSERT3U(sizeof (dnode_phys_t), ==, (1<<DNODE_SHIFT));
ASSERT((size & (sizeof (dnode_phys_t)-1)) == 0);
- size >>= DNODE_SHIFT;
- for (i = 0; i < size; i++) {
- dnode_byteswap(buf);
- buf++;
+ while (i < size) {
+ dnode_phys_t *dnp = (void *)(((char *)vbuf) + i);
+ dnode_byteswap(dnp);
+
+ i += DNODE_MIN_SIZE;
+ if (dnp->dn_type != DMU_OT_NONE)
+ i += dnp->dn_extra_slots * DNODE_MIN_SIZE;
}
}
void
dnode_setbonuslen(dnode_t *dn, int newsize, dmu_tx_t *tx)
{
- ASSERT3U(refcount_count(&dn->dn_holds), >=, 1);
+ ASSERT3U(zfs_refcount_count(&dn->dn_holds), >=, 1);
dnode_setdirty(dn, tx);
rw_enter(&dn->dn_struct_rwlock, RW_WRITER);
- ASSERT3U(newsize, <=, DN_MAX_BONUSLEN -
+ ASSERT3U(newsize, <=, DN_SLOTS_TO_BONUSLEN(dn->dn_num_slots) -
(dn->dn_nblkptr-1) * sizeof (blkptr_t));
+
+ if (newsize < dn->dn_bonuslen) {
+ /* clear any data after the end of the new size */
+ size_t diff = dn->dn_bonuslen - newsize;
+ char *data_end = ((char *)dn->dn_bonus->db.db_data) + newsize;
+ memset(data_end, 0, diff);
+ }
+
dn->dn_bonuslen = newsize;
if (newsize == 0)
dn->dn_next_bonuslen[tx->tx_txg & TXG_MASK] = DN_ZERO_BONUSLEN;
void
dnode_setbonus_type(dnode_t *dn, dmu_object_type_t newtype, dmu_tx_t *tx)
{
- ASSERT3U(refcount_count(&dn->dn_holds), >=, 1);
+ ASSERT3U(zfs_refcount_count(&dn->dn_holds), >=, 1);
dnode_setdirty(dn, tx);
rw_enter(&dn->dn_struct_rwlock, RW_WRITER);
dn->dn_bonustype = newtype;
void
dnode_rm_spill(dnode_t *dn, dmu_tx_t *tx)
{
- ASSERT3U(refcount_count(&dn->dn_holds), >=, 1);
+ ASSERT3U(zfs_refcount_count(&dn->dn_holds), >=, 1);
ASSERT(RW_WRITE_HELD(&dn->dn_struct_rwlock));
dnode_setdirty(dn, tx);
- dn->dn_rm_spillblk[tx->tx_txg&TXG_MASK] = DN_KILL_SPILLBLK;
+ dn->dn_rm_spillblk[tx->tx_txg & TXG_MASK] = DN_KILL_SPILLBLK;
dn->dn_have_spill = B_FALSE;
}
dnode_t *dn;
dn = kmem_cache_alloc(dnode_cache, KM_SLEEP);
- ASSERT(!POINTER_IS_VALID(dn->dn_objset));
dn->dn_moved = 0;
/*
dn->dn_compress = dnp->dn_compress;
dn->dn_bonustype = dnp->dn_bonustype;
dn->dn_bonuslen = dnp->dn_bonuslen;
+ dn->dn_num_slots = dnp->dn_extra_slots + 1;
dn->dn_maxblkid = dnp->dn_maxblkid;
dn->dn_have_spill = ((dnp->dn_flags & DNODE_FLAG_SPILL_BLKPTR) != 0);
dn->dn_id_flags = 0;
dmu_zfetch_init(&dn->dn_zfetch, dn);
ASSERT(DMU_OT_IS_VALID(dn->dn_phys->dn_type));
+ ASSERT(zrl_is_locked(&dnh->dnh_zrlock));
+ ASSERT(!DN_SLOT_IS_PTR(dnh->dnh_dnode));
mutex_enter(&os->os_lock);
- if (dnh->dnh_dnode != NULL) {
- /* Lost the allocation race. */
- mutex_exit(&os->os_lock);
- kmem_cache_free(dnode_cache, dn);
- return (dnh->dnh_dnode);
- }
/*
* Exclude special dnodes from os_dnodes so an empty os_dnodes
dnh->dnh_dnode = dn;
mutex_exit(&os->os_lock);
- arc_space_consume(sizeof (dnode_t), ARC_SPACE_OTHER);
+ arc_space_consume(sizeof (dnode_t), ARC_SPACE_DNODE);
+
return (dn);
}
mutex_exit(&os->os_lock);
/* the dnode can no longer move, so we can release the handle */
- zrl_remove(&dn->dn_handle->dnh_zrlock);
+ if (!zrl_is_locked(&dn->dn_handle->dnh_zrlock))
+ zrl_remove(&dn->dn_handle->dnh_zrlock);
dn->dn_allocated_txg = 0;
dn->dn_free_txg = 0;
dn->dn_assigned_txg = 0;
+ dn->dn_dirty_txg = 0;
dn->dn_dirtyctx = 0;
- if (dn->dn_dirtyctx_firstset != NULL) {
- kmem_free(dn->dn_dirtyctx_firstset, 1);
- dn->dn_dirtyctx_firstset = NULL;
- }
+ dn->dn_dirtyctx_firstset = NULL;
if (dn->dn_bonus != NULL) {
mutex_enter(&dn->dn_bonus->db_mtx);
- dbuf_evict(dn->dn_bonus);
+ dbuf_destroy(dn->dn_bonus);
dn->dn_bonus = NULL;
}
dn->dn_zio = NULL;
dn->dn_oldflags = 0;
dn->dn_olduid = 0;
dn->dn_oldgid = 0;
+ dn->dn_oldprojid = ZFS_DEFAULT_PROJID;
dn->dn_newuid = 0;
dn->dn_newgid = 0;
+ dn->dn_newprojid = ZFS_DEFAULT_PROJID;
dn->dn_id_flags = 0;
- dn->dn_unlisted_l0_blkid = 0;
- dmu_zfetch_rele(&dn->dn_zfetch);
+ dmu_zfetch_fini(&dn->dn_zfetch);
kmem_cache_free(dnode_cache, dn);
- arc_space_return(sizeof (dnode_t), ARC_SPACE_OTHER);
+ arc_space_return(sizeof (dnode_t), ARC_SPACE_DNODE);
if (complete_os_eviction)
dmu_objset_evict_done(os);
void
dnode_allocate(dnode_t *dn, dmu_object_type_t ot, int blocksize, int ibs,
- dmu_object_type_t bonustype, int bonuslen, dmu_tx_t *tx)
+ dmu_object_type_t bonustype, int bonuslen, int dn_slots, dmu_tx_t *tx)
{
int i;
+ ASSERT3U(dn_slots, >, 0);
+ ASSERT3U(dn_slots << DNODE_SHIFT, <=,
+ spa_maxdnodesize(dmu_objset_spa(dn->dn_objset)));
ASSERT3U(blocksize, <=,
spa_maxblocksize(dmu_objset_spa(dn->dn_objset)));
if (blocksize == 0)
ibs = MIN(MAX(ibs, DN_MIN_INDBLKSHIFT), DN_MAX_INDBLKSHIFT);
- dprintf("os=%p obj=%llu txg=%llu blocksize=%d ibs=%d\n", dn->dn_objset,
- dn->dn_object, tx->tx_txg, blocksize, ibs);
+ dprintf("os=%p obj=%llu txg=%llu blocksize=%d ibs=%d dn_slots=%d\n",
+ dn->dn_objset, (u_longlong_t)dn->dn_object,
+ (u_longlong_t)tx->tx_txg, blocksize, ibs, dn_slots);
+ DNODE_STAT_BUMP(dnode_allocate);
ASSERT(dn->dn_type == DMU_OT_NONE);
- ASSERT(bcmp(dn->dn_phys, &dnode_phys_zero, sizeof (dnode_phys_t)) == 0);
+ ASSERT0(memcmp(dn->dn_phys, &dnode_phys_zero, sizeof (dnode_phys_t)));
ASSERT(dn->dn_phys->dn_type == DMU_OT_NONE);
ASSERT(ot != DMU_OT_NONE);
ASSERT(DMU_OT_IS_VALID(ot));
ASSERT((bonustype == DMU_OT_NONE && bonuslen == 0) ||
(bonustype == DMU_OT_SA && bonuslen == 0) ||
+ (bonustype == DMU_OTN_UINT64_METADATA && bonuslen == 0) ||
(bonustype != DMU_OT_NONE && bonuslen != 0));
ASSERT(DMU_OT_IS_VALID(bonustype));
- ASSERT3U(bonuslen, <=, DN_MAX_BONUSLEN);
+ ASSERT3U(bonuslen, <=, DN_SLOTS_TO_BONUSLEN(dn_slots));
ASSERT(dn->dn_type == DMU_OT_NONE);
ASSERT0(dn->dn_maxblkid);
ASSERT0(dn->dn_allocated_txg);
ASSERT0(dn->dn_assigned_txg);
- ASSERT(refcount_is_zero(&dn->dn_tx_holds));
- ASSERT3U(refcount_count(&dn->dn_holds), <=, 1);
+ ASSERT(zfs_refcount_is_zero(&dn->dn_tx_holds));
+ ASSERT3U(zfs_refcount_count(&dn->dn_holds), <=, 1);
ASSERT(avl_is_empty(&dn->dn_dbufs));
for (i = 0; i < TXG_SIZE; i++) {
ASSERT0(dn->dn_next_bonustype[i]);
ASSERT0(dn->dn_rm_spillblk[i]);
ASSERT0(dn->dn_next_blksz[i]);
- ASSERT(!list_link_active(&dn->dn_dirty_link[i]));
+ ASSERT0(dn->dn_next_maxblkid[i]);
+ ASSERT(!multilist_link_active(&dn->dn_dirty_link[i]));
ASSERT3P(list_head(&dn->dn_dirty_records[i]), ==, NULL);
ASSERT3P(dn->dn_free_ranges[i], ==, NULL);
}
dnode_setdblksz(dn, blocksize);
dn->dn_indblkshift = ibs;
dn->dn_nlevels = 1;
+ dn->dn_num_slots = dn_slots;
if (bonustype == DMU_OT_SA) /* Maximize bonus space for SA */
dn->dn_nblkptr = 1;
- else
- dn->dn_nblkptr = 1 +
- ((DN_MAX_BONUSLEN - bonuslen) >> SPA_BLKPTRSHIFT);
+ else {
+ dn->dn_nblkptr = MIN(DN_MAX_NBLKPTR,
+ 1 + ((DN_SLOTS_TO_BONUSLEN(dn_slots) - bonuslen) >>
+ SPA_BLKPTRSHIFT));
+ }
+
dn->dn_bonustype = bonustype;
dn->dn_bonuslen = bonuslen;
dn->dn_checksum = ZIO_CHECKSUM_INHERIT;
dn->dn_dirtyctx = 0;
dn->dn_free_txg = 0;
- if (dn->dn_dirtyctx_firstset) {
- kmem_free(dn->dn_dirtyctx_firstset, 1);
- dn->dn_dirtyctx_firstset = NULL;
- }
+ dn->dn_dirtyctx_firstset = NULL;
+ dn->dn_dirty_txg = 0;
dn->dn_allocated_txg = tx->tx_txg;
dn->dn_id_flags = 0;
void
dnode_reallocate(dnode_t *dn, dmu_object_type_t ot, int blocksize,
- dmu_object_type_t bonustype, int bonuslen, dmu_tx_t *tx)
+ dmu_object_type_t bonustype, int bonuslen, int dn_slots,
+ boolean_t keep_spill, dmu_tx_t *tx)
{
int nblkptr;
(bonustype != DMU_OT_NONE && bonuslen != 0) ||
(bonustype == DMU_OT_SA && bonuslen == 0));
ASSERT(DMU_OT_IS_VALID(bonustype));
- ASSERT3U(bonuslen, <=, DN_MAX_BONUSLEN);
+ ASSERT3U(bonuslen, <=,
+ DN_BONUS_SIZE(spa_maxdnodesize(dmu_objset_spa(dn->dn_objset))));
+ ASSERT3U(bonuslen, <=, DN_BONUS_SIZE(dn_slots << DNODE_SHIFT));
+
+ dnode_free_interior_slots(dn);
+ DNODE_STAT_BUMP(dnode_reallocate);
/* clean up any unreferenced dbufs */
dnode_evict_dbufs(dn);
dnode_setdirty(dn, tx);
if (dn->dn_datablksz != blocksize) {
/* change blocksize */
- ASSERT(dn->dn_maxblkid == 0 &&
- (BP_IS_HOLE(&dn->dn_phys->dn_blkptr[0]) ||
- dnode_block_freed(dn, 0)));
+ ASSERT0(dn->dn_maxblkid);
+ ASSERT(BP_IS_HOLE(&dn->dn_phys->dn_blkptr[0]) ||
+ dnode_block_freed(dn, 0));
+
dnode_setdblksz(dn, blocksize);
- dn->dn_next_blksz[tx->tx_txg&TXG_MASK] = blocksize;
+ dn->dn_next_blksz[tx->tx_txg & TXG_MASK] = blocksize;
}
if (dn->dn_bonuslen != bonuslen)
- dn->dn_next_bonuslen[tx->tx_txg&TXG_MASK] = bonuslen;
+ dn->dn_next_bonuslen[tx->tx_txg & TXG_MASK] = bonuslen;
if (bonustype == DMU_OT_SA) /* Maximize bonus space for SA */
nblkptr = 1;
else
- nblkptr = 1 + ((DN_MAX_BONUSLEN - bonuslen) >> SPA_BLKPTRSHIFT);
+ nblkptr = MIN(DN_MAX_NBLKPTR,
+ 1 + ((DN_SLOTS_TO_BONUSLEN(dn_slots) - bonuslen) >>
+ SPA_BLKPTRSHIFT));
if (dn->dn_bonustype != bonustype)
- dn->dn_next_bonustype[tx->tx_txg&TXG_MASK] = bonustype;
+ dn->dn_next_bonustype[tx->tx_txg & TXG_MASK] = bonustype;
if (dn->dn_nblkptr != nblkptr)
- dn->dn_next_nblkptr[tx->tx_txg&TXG_MASK] = nblkptr;
- if (dn->dn_phys->dn_flags & DNODE_FLAG_SPILL_BLKPTR) {
+ dn->dn_next_nblkptr[tx->tx_txg & TXG_MASK] = nblkptr;
+ if (dn->dn_phys->dn_flags & DNODE_FLAG_SPILL_BLKPTR && !keep_spill) {
dbuf_rm_spill(dn, tx);
dnode_rm_spill(dn, tx);
}
+
rw_exit(&dn->dn_struct_rwlock);
/* change type */
mutex_enter(&dn->dn_mtx);
dn->dn_bonustype = bonustype;
dn->dn_bonuslen = bonuslen;
+ dn->dn_num_slots = dn_slots;
dn->dn_nblkptr = nblkptr;
dn->dn_checksum = ZIO_CHECKSUM_INHERIT;
dn->dn_compress = ZIO_COMPRESS_INHERIT;
/* fix up the bonus db_size */
if (dn->dn_bonus) {
dn->dn_bonus->db.db_size =
- DN_MAX_BONUSLEN - (dn->dn_nblkptr-1) * sizeof (blkptr_t);
+ DN_SLOTS_TO_BONUSLEN(dn->dn_num_slots) -
+ (dn->dn_nblkptr-1) * sizeof (blkptr_t);
ASSERT(dn->dn_bonuslen <= dn->dn_bonus->db.db_size);
}
}
#ifdef _KERNEL
-#ifdef DNODE_STATS
-static struct {
- uint64_t dms_dnode_invalid;
- uint64_t dms_dnode_recheck1;
- uint64_t dms_dnode_recheck2;
- uint64_t dms_dnode_special;
- uint64_t dms_dnode_handle;
- uint64_t dms_dnode_rwlock;
- uint64_t dms_dnode_active;
-} dnode_move_stats;
-#endif /* DNODE_STATS */
-
static void
dnode_move_impl(dnode_t *odn, dnode_t *ndn)
{
- int i;
-
ASSERT(!RW_LOCK_HELD(&odn->dn_struct_rwlock));
ASSERT(MUTEX_NOT_HELD(&odn->dn_mtx));
ASSERT(MUTEX_NOT_HELD(&odn->dn_dbufs_mtx));
- ASSERT(!RW_LOCK_HELD(&odn->dn_zfetch.zf_rwlock));
/* Copy fields. */
ndn->dn_objset = odn->dn_objset;
ndn->dn_datablkszsec = odn->dn_datablkszsec;
ndn->dn_datablksz = odn->dn_datablksz;
ndn->dn_maxblkid = odn->dn_maxblkid;
- bcopy(&odn->dn_next_nblkptr[0], &ndn->dn_next_nblkptr[0],
+ ndn->dn_num_slots = odn->dn_num_slots;
+ memcpy(ndn->dn_next_type, odn->dn_next_type,
+ sizeof (odn->dn_next_type));
+ memcpy(ndn->dn_next_nblkptr, odn->dn_next_nblkptr,
sizeof (odn->dn_next_nblkptr));
- bcopy(&odn->dn_next_nlevels[0], &ndn->dn_next_nlevels[0],
+ memcpy(ndn->dn_next_nlevels, odn->dn_next_nlevels,
sizeof (odn->dn_next_nlevels));
- bcopy(&odn->dn_next_indblkshift[0], &ndn->dn_next_indblkshift[0],
+ memcpy(ndn->dn_next_indblkshift, odn->dn_next_indblkshift,
sizeof (odn->dn_next_indblkshift));
- bcopy(&odn->dn_next_bonustype[0], &ndn->dn_next_bonustype[0],
+ memcpy(ndn->dn_next_bonustype, odn->dn_next_bonustype,
sizeof (odn->dn_next_bonustype));
- bcopy(&odn->dn_rm_spillblk[0], &ndn->dn_rm_spillblk[0],
+ memcpy(ndn->dn_rm_spillblk, odn->dn_rm_spillblk,
sizeof (odn->dn_rm_spillblk));
- bcopy(&odn->dn_next_bonuslen[0], &ndn->dn_next_bonuslen[0],
+ memcpy(ndn->dn_next_bonuslen, odn->dn_next_bonuslen,
sizeof (odn->dn_next_bonuslen));
- bcopy(&odn->dn_next_blksz[0], &ndn->dn_next_blksz[0],
+ memcpy(ndn->dn_next_blksz, odn->dn_next_blksz,
sizeof (odn->dn_next_blksz));
- for (i = 0; i < TXG_SIZE; i++) {
+ memcpy(ndn->dn_next_maxblkid, odn->dn_next_maxblkid,
+ sizeof (odn->dn_next_maxblkid));
+ for (int i = 0; i < TXG_SIZE; i++) {
list_move_tail(&ndn->dn_dirty_records[i],
&odn->dn_dirty_records[i]);
}
- bcopy(&odn->dn_free_ranges[0], &ndn->dn_free_ranges[0],
+ memcpy(ndn->dn_free_ranges, odn->dn_free_ranges,
sizeof (odn->dn_free_ranges));
ndn->dn_allocated_txg = odn->dn_allocated_txg;
ndn->dn_free_txg = odn->dn_free_txg;
ndn->dn_assigned_txg = odn->dn_assigned_txg;
+ ndn->dn_dirty_txg = odn->dn_dirty_txg;
ndn->dn_dirtyctx = odn->dn_dirtyctx;
ndn->dn_dirtyctx_firstset = odn->dn_dirtyctx_firstset;
- ASSERT(refcount_count(&odn->dn_tx_holds) == 0);
- refcount_transfer(&ndn->dn_holds, &odn->dn_holds);
+ ASSERT(zfs_refcount_count(&odn->dn_tx_holds) == 0);
+ zfs_refcount_transfer(&ndn->dn_holds, &odn->dn_holds);
ASSERT(avl_is_empty(&ndn->dn_dbufs));
avl_swap(&ndn->dn_dbufs, &odn->dn_dbufs);
ndn->dn_dbufs_count = odn->dn_dbufs_count;
- ndn->dn_unlisted_l0_blkid = odn->dn_unlisted_l0_blkid;
ndn->dn_bonus = odn->dn_bonus;
ndn->dn_have_spill = odn->dn_have_spill;
ndn->dn_zio = odn->dn_zio;
ndn->dn_oldflags = odn->dn_oldflags;
ndn->dn_olduid = odn->dn_olduid;
ndn->dn_oldgid = odn->dn_oldgid;
+ ndn->dn_oldprojid = odn->dn_oldprojid;
ndn->dn_newuid = odn->dn_newuid;
ndn->dn_newgid = odn->dn_newgid;
+ ndn->dn_newprojid = odn->dn_newprojid;
ndn->dn_id_flags = odn->dn_id_flags;
- dmu_zfetch_init(&ndn->dn_zfetch, NULL);
- list_move_tail(&ndn->dn_zfetch.zf_stream, &odn->dn_zfetch.zf_stream);
- ndn->dn_zfetch.zf_dnode = odn->dn_zfetch.zf_dnode;
- ndn->dn_zfetch.zf_stream_cnt = odn->dn_zfetch.zf_stream_cnt;
- ndn->dn_zfetch.zf_alloc_fail = odn->dn_zfetch.zf_alloc_fail;
+ dmu_zfetch_init(&ndn->dn_zfetch, ndn);
/*
* Update back pointers. Updating the handle fixes the back pointer of
*/
ASSERT(ndn->dn_handle->dnh_dnode == odn);
ndn->dn_handle->dnh_dnode = ndn;
- if (ndn->dn_zfetch.zf_dnode == odn) {
- ndn->dn_zfetch.zf_dnode = ndn;
- }
/*
* Invalidate the original dnode by clearing all of its back pointers.
avl_create(&odn->dn_dbufs, dbuf_compare, sizeof (dmu_buf_impl_t),
offsetof(dmu_buf_impl_t, db_link));
odn->dn_dbufs_count = 0;
- odn->dn_unlisted_l0_blkid = 0;
odn->dn_bonus = NULL;
- odn->dn_zfetch.zf_dnode = NULL;
+ dmu_zfetch_fini(&odn->dn_zfetch);
/*
* Set the low bit of the objset pointer to ensure that dnode_move()
/*
* Satisfy the destructor.
*/
- for (i = 0; i < TXG_SIZE; i++) {
+ for (int i = 0; i < TXG_SIZE; i++) {
list_create(&odn->dn_dirty_records[i],
sizeof (dbuf_dirty_record_t),
offsetof(dbuf_dirty_record_t, dr_dirty_node));
odn->dn_allocated_txg = 0;
odn->dn_free_txg = 0;
odn->dn_assigned_txg = 0;
+ odn->dn_dirty_txg = 0;
odn->dn_dirtyctx = 0;
odn->dn_dirtyctx_firstset = NULL;
odn->dn_have_spill = B_FALSE;
odn->dn_oldflags = 0;
odn->dn_olduid = 0;
odn->dn_oldgid = 0;
+ odn->dn_oldprojid = ZFS_DEFAULT_PROJID;
odn->dn_newuid = 0;
odn->dn_newgid = 0;
+ odn->dn_newprojid = ZFS_DEFAULT_PROJID;
odn->dn_id_flags = 0;
/*
odn->dn_moved = (uint8_t)-1;
}
-/*ARGSUSED*/
static kmem_cbrc_t
dnode_move(void *buf, void *newbuf, size_t size, void *arg)
{
*/
os = odn->dn_objset;
if (!POINTER_IS_VALID(os)) {
- DNODE_STAT_ADD(dnode_move_stats.dms_dnode_invalid);
+ DNODE_STAT_BUMP(dnode_move_invalid);
return (KMEM_CBRC_DONT_KNOW);
}
rw_enter(&os_lock, RW_WRITER);
if (os != odn->dn_objset) {
rw_exit(&os_lock);
- DNODE_STAT_ADD(dnode_move_stats.dms_dnode_recheck1);
+ DNODE_STAT_BUMP(dnode_move_recheck1);
return (KMEM_CBRC_DONT_KNOW);
}
if (os != odn->dn_objset) {
mutex_exit(&os->os_lock);
rw_exit(&os_lock);
- DNODE_STAT_ADD(dnode_move_stats.dms_dnode_recheck2);
+ DNODE_STAT_BUMP(dnode_move_recheck2);
return (KMEM_CBRC_DONT_KNOW);
}
rw_exit(&os_lock);
if (DMU_OBJECT_IS_SPECIAL(odn->dn_object)) {
mutex_exit(&os->os_lock);
- DNODE_STAT_ADD(dnode_move_stats.dms_dnode_special);
+ DNODE_STAT_BUMP(dnode_move_special);
return (KMEM_CBRC_NO);
}
ASSERT(odn->dn_dbuf != NULL); /* only "special" dnodes have no parent */
*/
if (!zrl_tryenter(&odn->dn_handle->dnh_zrlock)) {
mutex_exit(&os->os_lock);
- DNODE_STAT_ADD(dnode_move_stats.dms_dnode_handle);
+ DNODE_STAT_BUMP(dnode_move_handle);
return (KMEM_CBRC_LATER);
}
if (!rw_tryenter(&odn->dn_struct_rwlock, RW_WRITER)) {
zrl_exit(&odn->dn_handle->dnh_zrlock);
mutex_exit(&os->os_lock);
- DNODE_STAT_ADD(dnode_move_stats.dms_dnode_rwlock);
+ DNODE_STAT_BUMP(dnode_move_rwlock);
return (KMEM_CBRC_LATER);
}
* hold before the dbuf is removed, the hold is discounted, and the
* removal is blocked until the move completes.
*/
- refcount = refcount_count(&odn->dn_holds);
+ refcount = zfs_refcount_count(&odn->dn_holds);
ASSERT(refcount >= 0);
- dbufs = odn->dn_dbufs_count;
+ dbufs = DN_DBUFS_COUNT(odn);
/* We can't have more dbufs than dnode holds. */
ASSERT3U(dbufs, <=, refcount);
rw_exit(&odn->dn_struct_rwlock);
zrl_exit(&odn->dn_handle->dnh_zrlock);
mutex_exit(&os->os_lock);
- DNODE_STAT_ADD(dnode_move_stats.dms_dnode_active);
+ DNODE_STAT_BUMP(dnode_move_active);
return (KMEM_CBRC_LATER);
}
list_link_replace(&odn->dn_link, &ndn->dn_link);
/* If the dnode was safe to move, the refcount cannot have changed. */
- ASSERT(refcount == refcount_count(&ndn->dn_holds));
- ASSERT(dbufs == ndn->dn_dbufs_count);
+ ASSERT(refcount == zfs_refcount_count(&ndn->dn_holds));
+ ASSERT(dbufs == DN_DBUFS_COUNT(ndn));
zrl_exit(&ndn->dn_handle->dnh_zrlock); /* handle has moved */
mutex_exit(&os->os_lock);
}
#endif /* _KERNEL */
+static void
+dnode_slots_hold(dnode_children_t *children, int idx, int slots)
+{
+ ASSERT3S(idx + slots, <=, DNODES_PER_BLOCK);
+
+ for (int i = idx; i < idx + slots; i++) {
+ dnode_handle_t *dnh = &children->dnc_children[i];
+ zrl_add(&dnh->dnh_zrlock);
+ }
+}
+
+static void
+dnode_slots_rele(dnode_children_t *children, int idx, int slots)
+{
+ ASSERT3S(idx + slots, <=, DNODES_PER_BLOCK);
+
+ for (int i = idx; i < idx + slots; i++) {
+ dnode_handle_t *dnh = &children->dnc_children[i];
+
+ if (zrl_is_locked(&dnh->dnh_zrlock))
+ zrl_exit(&dnh->dnh_zrlock);
+ else
+ zrl_remove(&dnh->dnh_zrlock);
+ }
+}
+
+static int
+dnode_slots_tryenter(dnode_children_t *children, int idx, int slots)
+{
+ ASSERT3S(idx + slots, <=, DNODES_PER_BLOCK);
+
+ for (int i = idx; i < idx + slots; i++) {
+ dnode_handle_t *dnh = &children->dnc_children[i];
+
+ if (!zrl_tryenter(&dnh->dnh_zrlock)) {
+ for (int j = idx; j < i; j++) {
+ dnh = &children->dnc_children[j];
+ zrl_exit(&dnh->dnh_zrlock);
+ }
+
+ return (0);
+ }
+ }
+
+ return (1);
+}
+
+static void
+dnode_set_slots(dnode_children_t *children, int idx, int slots, void *ptr)
+{
+ ASSERT3S(idx + slots, <=, DNODES_PER_BLOCK);
+
+ for (int i = idx; i < idx + slots; i++) {
+ dnode_handle_t *dnh = &children->dnc_children[i];
+ dnh->dnh_dnode = ptr;
+ }
+}
+
+static boolean_t
+dnode_check_slots_free(dnode_children_t *children, int idx, int slots)
+{
+ ASSERT3S(idx + slots, <=, DNODES_PER_BLOCK);
+
+ /*
+ * If all dnode slots are either already free or
+ * evictable return B_TRUE.
+ */
+ for (int i = idx; i < idx + slots; i++) {
+ dnode_handle_t *dnh = &children->dnc_children[i];
+ dnode_t *dn = dnh->dnh_dnode;
+
+ if (dn == DN_SLOT_FREE) {
+ continue;
+ } else if (DN_SLOT_IS_PTR(dn)) {
+ mutex_enter(&dn->dn_mtx);
+ boolean_t can_free = (dn->dn_type == DMU_OT_NONE &&
+ zfs_refcount_is_zero(&dn->dn_holds) &&
+ !DNODE_IS_DIRTY(dn));
+ mutex_exit(&dn->dn_mtx);
+
+ if (!can_free)
+ return (B_FALSE);
+ else
+ continue;
+ } else {
+ return (B_FALSE);
+ }
+ }
+
+ return (B_TRUE);
+}
+
+static uint_t
+dnode_reclaim_slots(dnode_children_t *children, int idx, int slots)
+{
+ uint_t reclaimed = 0;
+
+ ASSERT3S(idx + slots, <=, DNODES_PER_BLOCK);
+
+ for (int i = idx; i < idx + slots; i++) {
+ dnode_handle_t *dnh = &children->dnc_children[i];
+
+ ASSERT(zrl_is_locked(&dnh->dnh_zrlock));
+
+ if (DN_SLOT_IS_PTR(dnh->dnh_dnode)) {
+ ASSERT3S(dnh->dnh_dnode->dn_type, ==, DMU_OT_NONE);
+ dnode_destroy(dnh->dnh_dnode);
+ dnh->dnh_dnode = DN_SLOT_FREE;
+ reclaimed++;
+ }
+ }
+
+ return (reclaimed);
+}
+
+void
+dnode_free_interior_slots(dnode_t *dn)
+{
+ dnode_children_t *children = dmu_buf_get_user(&dn->dn_dbuf->db);
+ int epb = dn->dn_dbuf->db.db_size >> DNODE_SHIFT;
+ int idx = (dn->dn_object & (epb - 1)) + 1;
+ int slots = dn->dn_num_slots - 1;
+
+ if (slots == 0)
+ return;
+
+ ASSERT3S(idx + slots, <=, DNODES_PER_BLOCK);
+
+ while (!dnode_slots_tryenter(children, idx, slots)) {
+ DNODE_STAT_BUMP(dnode_free_interior_lock_retry);
+ kpreempt(KPREEMPT_SYNC);
+ }
+
+ dnode_set_slots(children, idx, slots, DN_SLOT_FREE);
+ dnode_slots_rele(children, idx, slots);
+}
+
void
dnode_special_close(dnode_handle_t *dnh)
{
dnode_t *dn = dnh->dnh_dnode;
/*
- * Wait for final references to the dnode to clear. This can
- * only happen if the arc is asyncronously evicting state that
- * has a hold on this dnode while we are trying to evict this
- * dnode.
+ * Ensure dnode_rele_and_unlock() has released dn_mtx, after final
+ * zfs_refcount_remove()
*/
- while (refcount_count(&dn->dn_holds) > 0)
- delay(1);
+ mutex_enter(&dn->dn_mtx);
+ if (zfs_refcount_count(&dn->dn_holds) > 0)
+ cv_wait(&dn->dn_nodnholds, &dn->dn_mtx);
+ mutex_exit(&dn->dn_mtx);
+ ASSERT3U(zfs_refcount_count(&dn->dn_holds), ==, 0);
+
ASSERT(dn->dn_dbuf == NULL ||
dmu_buf_get_user(&dn->dn_dbuf->db) == NULL);
zrl_add(&dnh->dnh_zrlock);
{
dnode_t *dn;
- dn = dnode_create(os, dnp, NULL, object, dnh);
zrl_init(&dnh->dnh_zrlock);
+ VERIFY3U(1, ==, zrl_tryenter(&dnh->dnh_zrlock));
+
+ dn = dnode_create(os, dnp, NULL, object, dnh);
DNODE_VERIFY(dn);
+
+ zrl_exit(&dnh->dnh_zrlock);
}
static void
-dnode_buf_pageout(void *dbu)
+dnode_buf_evict_async(void *dbu)
{
- dnode_children_t *children_dnodes = dbu;
- int i;
+ dnode_children_t *dnc = dbu;
+
+ DNODE_STAT_BUMP(dnode_buf_evict);
- for (i = 0; i < children_dnodes->dnc_count; i++) {
- dnode_handle_t *dnh = &children_dnodes->dnc_children[i];
+ for (int i = 0; i < dnc->dnc_count; i++) {
+ dnode_handle_t *dnh = &dnc->dnc_children[i];
dnode_t *dn;
/*
* another valid address, so there is no need here to guard
* against changes to or from NULL.
*/
- if (dnh->dnh_dnode == NULL) {
+ if (!DN_SLOT_IS_PTR(dnh->dnh_dnode)) {
zrl_destroy(&dnh->dnh_zrlock);
+ dnh->dnh_dnode = DN_SLOT_UNINIT;
continue;
}
* it wouldn't be eligible for eviction and this function
* would not have been called.
*/
- ASSERT(refcount_is_zero(&dn->dn_holds));
- ASSERT(refcount_is_zero(&dn->dn_tx_holds));
+ ASSERT(zfs_refcount_is_zero(&dn->dn_holds));
+ ASSERT(zfs_refcount_is_zero(&dn->dn_tx_holds));
- dnode_destroy(dn); /* implicit zrl_remove() */
+ dnode_destroy(dn); /* implicit zrl_remove() for first slot */
zrl_destroy(&dnh->dnh_zrlock);
- dnh->dnh_dnode = NULL;
+ dnh->dnh_dnode = DN_SLOT_UNINIT;
}
- kmem_free(children_dnodes, sizeof (dnode_children_t) +
- children_dnodes->dnc_count * sizeof (dnode_handle_t));
+ kmem_free(dnc, sizeof (dnode_children_t) +
+ dnc->dnc_count * sizeof (dnode_handle_t));
}
/*
+ * When the DNODE_MUST_BE_FREE flag is set, the "slots" parameter is used
+ * to ensure the hole at the specified object offset is large enough to
+ * hold the dnode being created. The slots parameter is also used to ensure
+ * a dnode does not span multiple dnode blocks. In both of these cases, if
+ * a failure occurs, ENOSPC is returned. Keep in mind, these failure cases
+ * are only possible when using DNODE_MUST_BE_FREE.
+ *
+ * If the DNODE_MUST_BE_ALLOCATED flag is set, "slots" must be 0.
+ * dnode_hold_impl() will check if the requested dnode is already consumed
+ * as an extra dnode slot by an large dnode, in which case it returns
+ * ENOENT.
+ *
+ * If the DNODE_DRY_RUN flag is set, we don't actually hold the dnode, just
+ * return whether the hold would succeed or not. tag and dnp should set to
+ * NULL in this case.
+ *
* errors:
- * EINVAL - invalid object number.
- * EIO - i/o error.
+ * EINVAL - Invalid object number or flags.
+ * ENOSPC - Hole too small to fulfill "slots" request (DNODE_MUST_BE_FREE)
+ * EEXIST - Refers to an allocated dnode (DNODE_MUST_BE_FREE)
+ * - Refers to a freeing dnode (DNODE_MUST_BE_FREE)
+ * - Refers to an interior dnode slot (DNODE_MUST_BE_ALLOCATED)
+ * ENOENT - The requested dnode is not allocated (DNODE_MUST_BE_ALLOCATED)
+ * - The requested dnode is being freed (DNODE_MUST_BE_ALLOCATED)
+ * EIO - I/O error when reading the meta dnode dbuf.
+ *
* succeeds even for free dnodes.
*/
int
-dnode_hold_impl(objset_t *os, uint64_t object, int flag,
- void *tag, dnode_t **dnp)
+dnode_hold_impl(objset_t *os, uint64_t object, int flag, int slots,
+ const void *tag, dnode_t **dnp)
{
int epb, idx, err;
int drop_struct_lock = FALSE;
uint64_t blk;
dnode_t *mdn, *dn;
dmu_buf_impl_t *db;
- dnode_children_t *children_dnodes;
+ dnode_children_t *dnc;
+ dnode_phys_t *dn_block;
dnode_handle_t *dnh;
+ ASSERT(!(flag & DNODE_MUST_BE_ALLOCATED) || (slots == 0));
+ ASSERT(!(flag & DNODE_MUST_BE_FREE) || (slots > 0));
+ IMPLY(flag & DNODE_DRY_RUN, (tag == NULL) && (dnp == NULL));
+
/*
* If you are holding the spa config lock as writer, you shouldn't
* be asking the DMU to do *anything* unless it's the root pool
(spa_is_root(os->os_spa) &&
spa_config_held(os->os_spa, SCL_STATE, RW_WRITER)));
- if (object == DMU_USERUSED_OBJECT || object == DMU_GROUPUSED_OBJECT) {
- dn = (object == DMU_USERUSED_OBJECT) ?
- DMU_USERUSED_DNODE(os) : DMU_GROUPUSED_DNODE(os);
+ ASSERT((flag & DNODE_MUST_BE_ALLOCATED) || (flag & DNODE_MUST_BE_FREE));
+
+ if (object == DMU_USERUSED_OBJECT || object == DMU_GROUPUSED_OBJECT ||
+ object == DMU_PROJECTUSED_OBJECT) {
+ if (object == DMU_USERUSED_OBJECT)
+ dn = DMU_USERUSED_DNODE(os);
+ else if (object == DMU_GROUPUSED_OBJECT)
+ dn = DMU_GROUPUSED_DNODE(os);
+ else
+ dn = DMU_PROJECTUSED_DNODE(os);
if (dn == NULL)
return (SET_ERROR(ENOENT));
type = dn->dn_type;
if ((flag & DNODE_MUST_BE_FREE) && type != DMU_OT_NONE)
return (SET_ERROR(EEXIST));
DNODE_VERIFY(dn);
- (void) refcount_add(&dn->dn_holds, tag);
- *dnp = dn;
+ /* Don't actually hold if dry run, just return 0 */
+ if (!(flag & DNODE_DRY_RUN)) {
+ (void) zfs_refcount_add(&dn->dn_holds, tag);
+ *dnp = dn;
+ }
return (0);
}
drop_struct_lock = TRUE;
}
- blk = dbuf_whichblock(mdn, object * sizeof (dnode_phys_t));
-
+ blk = dbuf_whichblock(mdn, 0, object * sizeof (dnode_phys_t));
db = dbuf_hold(mdn, blk, FTAG);
if (drop_struct_lock)
rw_exit(&mdn->dn_struct_rwlock);
- if (db == NULL)
+ if (db == NULL) {
+ DNODE_STAT_BUMP(dnode_hold_dbuf_hold);
return (SET_ERROR(EIO));
- err = dbuf_read(db, NULL, DB_RF_CANFAIL);
+ }
+
+ /*
+ * We do not need to decrypt to read the dnode so it doesn't matter
+ * if we get the encrypted or decrypted version.
+ */
+ err = dbuf_read(db, NULL, DB_RF_CANFAIL |
+ DB_RF_NO_DECRYPT | DB_RF_NOPREFETCH);
if (err) {
+ DNODE_STAT_BUMP(dnode_hold_dbuf_read);
dbuf_rele(db, FTAG);
return (err);
}
ASSERT3U(db->db.db_size, >=, 1<<DNODE_SHIFT);
epb = db->db.db_size >> DNODE_SHIFT;
- idx = object & (epb-1);
+ idx = object & (epb - 1);
+ dn_block = (dnode_phys_t *)db->db.db_data;
ASSERT(DB_DNODE(db)->dn_type == DMU_OT_DNODE);
- children_dnodes = dmu_buf_get_user(&db->db);
- if (children_dnodes == NULL) {
- int i;
+ dnc = dmu_buf_get_user(&db->db);
+ dnh = NULL;
+ if (dnc == NULL) {
dnode_children_t *winner;
- children_dnodes = kmem_zalloc(sizeof (dnode_children_t) +
+ int skip = 0;
+
+ dnc = kmem_zalloc(sizeof (dnode_children_t) +
epb * sizeof (dnode_handle_t), KM_SLEEP);
- children_dnodes->dnc_count = epb;
- dnh = &children_dnodes->dnc_children[0];
- for (i = 0; i < epb; i++) {
+ dnc->dnc_count = epb;
+ dnh = &dnc->dnc_children[0];
+
+ /* Initialize dnode slot status from dnode_phys_t */
+ for (int i = 0; i < epb; i++) {
zrl_init(&dnh[i].dnh_zrlock);
+
+ if (skip) {
+ skip--;
+ continue;
+ }
+
+ if (dn_block[i].dn_type != DMU_OT_NONE) {
+ int interior = dn_block[i].dn_extra_slots;
+
+ dnode_set_slots(dnc, i, 1, DN_SLOT_ALLOCATED);
+ dnode_set_slots(dnc, i + 1, interior,
+ DN_SLOT_INTERIOR);
+ skip = interior;
+ } else {
+ dnh[i].dnh_dnode = DN_SLOT_FREE;
+ skip = 0;
+ }
}
- dmu_buf_init_user(&children_dnodes->dnc_dbu,
- dnode_buf_pageout, NULL);
- winner = dmu_buf_set_user(&db->db, &children_dnodes->dnc_dbu);
+
+ dmu_buf_init_user(&dnc->dnc_dbu, NULL,
+ dnode_buf_evict_async, NULL);
+ winner = dmu_buf_set_user(&db->db, &dnc->dnc_dbu);
if (winner != NULL) {
- for (i = 0; i < epb; i++) {
+ for (int i = 0; i < epb; i++)
zrl_destroy(&dnh[i].dnh_zrlock);
- }
- kmem_free(children_dnodes, sizeof (dnode_children_t) +
+ kmem_free(dnc, sizeof (dnode_children_t) +
epb * sizeof (dnode_handle_t));
- children_dnodes = winner;
+ dnc = winner;
}
}
- ASSERT(children_dnodes->dnc_count == epb);
- dnh = &children_dnodes->dnc_children[idx];
- zrl_add(&dnh->dnh_zrlock);
- dn = dnh->dnh_dnode;
- if (dn == NULL) {
- dnode_phys_t *phys = (dnode_phys_t *)db->db.db_data+idx;
+ ASSERT(dnc->dnc_count == epb);
- dn = dnode_create(os, phys, db, object, dnh);
- }
+ if (flag & DNODE_MUST_BE_ALLOCATED) {
+ slots = 1;
- mutex_enter(&dn->dn_mtx);
- type = dn->dn_type;
- if (dn->dn_free_txg ||
- ((flag & DNODE_MUST_BE_ALLOCATED) && type == DMU_OT_NONE) ||
- ((flag & DNODE_MUST_BE_FREE) &&
- (type != DMU_OT_NONE || !refcount_is_zero(&dn->dn_holds)))) {
- mutex_exit(&dn->dn_mtx);
- zrl_remove(&dnh->dnh_zrlock);
+ dnode_slots_hold(dnc, idx, slots);
+ dnh = &dnc->dnc_children[idx];
+
+ if (DN_SLOT_IS_PTR(dnh->dnh_dnode)) {
+ dn = dnh->dnh_dnode;
+ } else if (dnh->dnh_dnode == DN_SLOT_INTERIOR) {
+ DNODE_STAT_BUMP(dnode_hold_alloc_interior);
+ dnode_slots_rele(dnc, idx, slots);
+ dbuf_rele(db, FTAG);
+ return (SET_ERROR(EEXIST));
+ } else if (dnh->dnh_dnode != DN_SLOT_ALLOCATED) {
+ DNODE_STAT_BUMP(dnode_hold_alloc_misses);
+ dnode_slots_rele(dnc, idx, slots);
+ dbuf_rele(db, FTAG);
+ return (SET_ERROR(ENOENT));
+ } else {
+ dnode_slots_rele(dnc, idx, slots);
+ while (!dnode_slots_tryenter(dnc, idx, slots)) {
+ DNODE_STAT_BUMP(dnode_hold_alloc_lock_retry);
+ kpreempt(KPREEMPT_SYNC);
+ }
+
+ /*
+ * Someone else won the race and called dnode_create()
+ * after we checked DN_SLOT_IS_PTR() above but before
+ * we acquired the lock.
+ */
+ if (DN_SLOT_IS_PTR(dnh->dnh_dnode)) {
+ DNODE_STAT_BUMP(dnode_hold_alloc_lock_misses);
+ dn = dnh->dnh_dnode;
+ } else {
+ dn = dnode_create(os, dn_block + idx, db,
+ object, dnh);
+ dmu_buf_add_user_size(&db->db,
+ sizeof (dnode_t));
+ }
+ }
+
+ mutex_enter(&dn->dn_mtx);
+ if (dn->dn_type == DMU_OT_NONE || dn->dn_free_txg != 0) {
+ DNODE_STAT_BUMP(dnode_hold_alloc_type_none);
+ mutex_exit(&dn->dn_mtx);
+ dnode_slots_rele(dnc, idx, slots);
+ dbuf_rele(db, FTAG);
+ return (SET_ERROR(ENOENT));
+ }
+
+ /* Don't actually hold if dry run, just return 0 */
+ if (flag & DNODE_DRY_RUN) {
+ mutex_exit(&dn->dn_mtx);
+ dnode_slots_rele(dnc, idx, slots);
+ dbuf_rele(db, FTAG);
+ return (0);
+ }
+
+ DNODE_STAT_BUMP(dnode_hold_alloc_hits);
+ } else if (flag & DNODE_MUST_BE_FREE) {
+
+ if (idx + slots - 1 >= DNODES_PER_BLOCK) {
+ DNODE_STAT_BUMP(dnode_hold_free_overflow);
+ dbuf_rele(db, FTAG);
+ return (SET_ERROR(ENOSPC));
+ }
+
+ dnode_slots_hold(dnc, idx, slots);
+
+ if (!dnode_check_slots_free(dnc, idx, slots)) {
+ DNODE_STAT_BUMP(dnode_hold_free_misses);
+ dnode_slots_rele(dnc, idx, slots);
+ dbuf_rele(db, FTAG);
+ return (SET_ERROR(ENOSPC));
+ }
+
+ dnode_slots_rele(dnc, idx, slots);
+ while (!dnode_slots_tryenter(dnc, idx, slots)) {
+ DNODE_STAT_BUMP(dnode_hold_free_lock_retry);
+ kpreempt(KPREEMPT_SYNC);
+ }
+
+ if (!dnode_check_slots_free(dnc, idx, slots)) {
+ DNODE_STAT_BUMP(dnode_hold_free_lock_misses);
+ dnode_slots_rele(dnc, idx, slots);
+ dbuf_rele(db, FTAG);
+ return (SET_ERROR(ENOSPC));
+ }
+
+ /*
+ * Allocated but otherwise free dnodes which would
+ * be in the interior of a multi-slot dnodes need
+ * to be freed. Single slot dnodes can be safely
+ * re-purposed as a performance optimization.
+ */
+ if (slots > 1) {
+ uint_t reclaimed =
+ dnode_reclaim_slots(dnc, idx + 1, slots - 1);
+ if (reclaimed > 0)
+ dmu_buf_sub_user_size(&db->db,
+ reclaimed * sizeof (dnode_t));
+ }
+
+ dnh = &dnc->dnc_children[idx];
+ if (DN_SLOT_IS_PTR(dnh->dnh_dnode)) {
+ dn = dnh->dnh_dnode;
+ } else {
+ dn = dnode_create(os, dn_block + idx, db,
+ object, dnh);
+ dmu_buf_add_user_size(&db->db, sizeof (dnode_t));
+ }
+
+ mutex_enter(&dn->dn_mtx);
+ if (!zfs_refcount_is_zero(&dn->dn_holds) || dn->dn_free_txg) {
+ DNODE_STAT_BUMP(dnode_hold_free_refcount);
+ mutex_exit(&dn->dn_mtx);
+ dnode_slots_rele(dnc, idx, slots);
+ dbuf_rele(db, FTAG);
+ return (SET_ERROR(EEXIST));
+ }
+
+ /* Don't actually hold if dry run, just return 0 */
+ if (flag & DNODE_DRY_RUN) {
+ mutex_exit(&dn->dn_mtx);
+ dnode_slots_rele(dnc, idx, slots);
+ dbuf_rele(db, FTAG);
+ return (0);
+ }
+
+ dnode_set_slots(dnc, idx + 1, slots - 1, DN_SLOT_INTERIOR);
+ DNODE_STAT_BUMP(dnode_hold_free_hits);
+ } else {
dbuf_rele(db, FTAG);
- return (type == DMU_OT_NONE ? ENOENT : EEXIST);
+ return (SET_ERROR(EINVAL));
}
- if (refcount_add(&dn->dn_holds, tag) == 1)
+
+ ASSERT0(dn->dn_free_txg);
+
+ if (zfs_refcount_add(&dn->dn_holds, tag) == 1)
dbuf_add_ref(db, dnh);
+
mutex_exit(&dn->dn_mtx);
/* Now we can rely on the hold to prevent the dnode from moving. */
- zrl_remove(&dnh->dnh_zrlock);
+ dnode_slots_rele(dnc, idx, slots);
DNODE_VERIFY(dn);
+ ASSERT3P(dnp, !=, NULL);
ASSERT3P(dn->dn_dbuf, ==, db);
ASSERT3U(dn->dn_object, ==, object);
dbuf_rele(db, FTAG);
* Return held dnode if the object is allocated, NULL if not.
*/
int
-dnode_hold(objset_t *os, uint64_t object, void *tag, dnode_t **dnp)
+dnode_hold(objset_t *os, uint64_t object, const void *tag, dnode_t **dnp)
{
- return (dnode_hold_impl(os, object, DNODE_MUST_BE_ALLOCATED, tag, dnp));
+ return (dnode_hold_impl(os, object, DNODE_MUST_BE_ALLOCATED, 0, tag,
+ dnp));
}
/*
* new reference.
*/
boolean_t
-dnode_add_ref(dnode_t *dn, void *tag)
+dnode_add_ref(dnode_t *dn, const void *tag)
{
mutex_enter(&dn->dn_mtx);
- if (refcount_is_zero(&dn->dn_holds)) {
+ if (zfs_refcount_is_zero(&dn->dn_holds)) {
mutex_exit(&dn->dn_mtx);
return (FALSE);
}
- VERIFY(1 < refcount_add(&dn->dn_holds, tag));
+ VERIFY(1 < zfs_refcount_add(&dn->dn_holds, tag));
mutex_exit(&dn->dn_mtx);
return (TRUE);
}
void
-dnode_rele(dnode_t *dn, void *tag)
+dnode_rele(dnode_t *dn, const void *tag)
{
mutex_enter(&dn->dn_mtx);
- dnode_rele_and_unlock(dn, tag);
+ dnode_rele_and_unlock(dn, tag, B_FALSE);
}
void
-dnode_rele_and_unlock(dnode_t *dn, void *tag)
+dnode_rele_and_unlock(dnode_t *dn, const void *tag, boolean_t evicting)
{
uint64_t refs;
/* Get while the hold prevents the dnode from moving. */
dmu_buf_impl_t *db = dn->dn_dbuf;
dnode_handle_t *dnh = dn->dn_handle;
- refs = refcount_remove(&dn->dn_holds, tag);
+ refs = zfs_refcount_remove(&dn->dn_holds, tag);
+ if (refs == 0)
+ cv_broadcast(&dn->dn_nodnholds);
mutex_exit(&dn->dn_mtx);
+ /* dnode could get destroyed at this point, so don't use it anymore */
/*
* It's unsafe to release the last hold on a dnode by dnode_rele() or
* other direct or indirect hold on the dnode must first drop the dnode
* handle.
*/
+#ifdef ZFS_DEBUG
ASSERT(refs > 0 || dnh->dnh_zrlock.zr_owner != curthread);
+#endif
/* NOTE: the DNODE_DNODE does not have a dn_dbuf */
if (refs == 0 && db != NULL) {
* that the handle has zero references, but that will be
* asserted anyway when the handle gets destroyed.
*/
- dbuf_rele(db, dnh);
+ mutex_enter(&db->db_mtx);
+ dbuf_rele_and_unlock(db, dnh, evicting);
}
}
+/*
+ * Test whether we can create a dnode at the specified location.
+ */
+int
+dnode_try_claim(objset_t *os, uint64_t object, int slots)
+{
+ return (dnode_hold_impl(os, object, DNODE_MUST_BE_FREE | DNODE_DRY_RUN,
+ slots, NULL, NULL));
+}
+
+/*
+ * Checks if the dnode itself is dirty, or is carrying any uncommitted records.
+ * It is important to check both conditions, as some operations (eg appending
+ * to a file) can dirty both as a single logical unit, but they are not synced
+ * out atomically, so checking one and not the other can result in an object
+ * appearing to be clean mid-way through a commit.
+ *
+ * Do not change this lightly! If you get it wrong, dmu_offset_next() can
+ * detect a hole where there is really data, leading to silent corruption.
+ */
+boolean_t
+dnode_is_dirty(dnode_t *dn)
+{
+ mutex_enter(&dn->dn_mtx);
+
+ for (int i = 0; i < TXG_SIZE; i++) {
+ if (multilist_link_active(&dn->dn_dirty_link[i]) ||
+ !list_is_empty(&dn->dn_dirty_records[i])) {
+ mutex_exit(&dn->dn_mtx);
+ return (B_TRUE);
+ }
+ }
+
+ mutex_exit(&dn->dn_mtx);
+
+ return (B_FALSE);
+}
+
void
dnode_setdirty(dnode_t *dn, dmu_tx_t *tx)
{
*/
dmu_objset_userquota_get_ids(dn, B_TRUE, tx);
- mutex_enter(&os->os_lock);
+ multilist_t *dirtylist = &os->os_dirty_dnodes[txg & TXG_MASK];
+ multilist_sublist_t *mls = multilist_sublist_lock_obj(dirtylist, dn);
/*
* If we are already marked dirty, we're done.
*/
- if (list_link_active(&dn->dn_dirty_link[txg & TXG_MASK])) {
- mutex_exit(&os->os_lock);
+ if (multilist_link_active(&dn->dn_dirty_link[txg & TXG_MASK])) {
+ multilist_sublist_unlock(mls);
return;
}
- ASSERT(!refcount_is_zero(&dn->dn_holds) ||
+ ASSERT(!zfs_refcount_is_zero(&dn->dn_holds) ||
!avl_is_empty(&dn->dn_dbufs));
ASSERT(dn->dn_datablksz != 0);
- ASSERT0(dn->dn_next_bonuslen[txg&TXG_MASK]);
- ASSERT0(dn->dn_next_blksz[txg&TXG_MASK]);
- ASSERT0(dn->dn_next_bonustype[txg&TXG_MASK]);
+ ASSERT0(dn->dn_next_bonuslen[txg & TXG_MASK]);
+ ASSERT0(dn->dn_next_blksz[txg & TXG_MASK]);
+ ASSERT0(dn->dn_next_bonustype[txg & TXG_MASK]);
dprintf_ds(os->os_dsl_dataset, "obj=%llu txg=%llu\n",
- dn->dn_object, txg);
+ (u_longlong_t)dn->dn_object, (u_longlong_t)txg);
- if (dn->dn_free_txg > 0 && dn->dn_free_txg <= txg) {
- list_insert_tail(&os->os_free_dnodes[txg&TXG_MASK], dn);
- } else {
- list_insert_tail(&os->os_dirty_dnodes[txg&TXG_MASK], dn);
- }
+ multilist_sublist_insert_head(mls, dn);
- mutex_exit(&os->os_lock);
+ multilist_sublist_unlock(mls);
/*
* The dnode maintains a hold on its containing dbuf as
void
dnode_free(dnode_t *dn, dmu_tx_t *tx)
{
- int txgoff = tx->tx_txg & TXG_MASK;
-
- dprintf("dn=%p txg=%llu\n", dn, tx->tx_txg);
-
- /* we should be the only holder... hopefully */
- /* ASSERT3U(refcount_count(&dn->dn_holds), ==, 1); */
-
mutex_enter(&dn->dn_mtx);
if (dn->dn_type == DMU_OT_NONE || dn->dn_free_txg) {
mutex_exit(&dn->dn_mtx);
dn->dn_free_txg = tx->tx_txg;
mutex_exit(&dn->dn_mtx);
- /*
- * If the dnode is already dirty, it needs to be moved from
- * the dirty list to the free list.
- */
- mutex_enter(&dn->dn_objset->os_lock);
- if (list_link_active(&dn->dn_dirty_link[txgoff])) {
- list_remove(&dn->dn_objset->os_dirty_dnodes[txgoff], dn);
- list_insert_tail(&dn->dn_objset->os_free_dnodes[txgoff], dn);
- mutex_exit(&dn->dn_objset->os_lock);
- } else {
- mutex_exit(&dn->dn_objset->os_lock);
- dnode_setdirty(dn, tx);
- }
+ dnode_setdirty(dn, tx);
}
/*
if (ibs == dn->dn_indblkshift)
ibs = 0;
- if (size >> SPA_MINBLOCKSHIFT == dn->dn_datablkszsec && ibs == 0)
+ if (size == dn->dn_datablksz && ibs == 0)
return (0);
rw_enter(&dn->dn_struct_rwlock, RW_WRITER);
if (ibs && dn->dn_nlevels != 1)
goto fail;
- /* resize the old block */
- err = dbuf_hold_impl(dn, 0, 0, TRUE, FTAG, &db);
- if (err == 0)
- dbuf_new_size(db, size, tx);
- else if (err != ENOENT)
- goto fail;
-
- dnode_setdblksz(dn, size);
dnode_setdirty(dn, tx);
- dn->dn_next_blksz[tx->tx_txg&TXG_MASK] = size;
+ if (size != dn->dn_datablksz) {
+ /* resize the old block */
+ err = dbuf_hold_impl(dn, 0, 0, TRUE, FALSE, FTAG, &db);
+ if (err == 0) {
+ dbuf_new_size(db, size, tx);
+ } else if (err != ENOENT) {
+ goto fail;
+ }
+
+ dnode_setdblksz(dn, size);
+ dn->dn_next_blksz[tx->tx_txg & TXG_MASK] = size;
+ if (db)
+ dbuf_rele(db, FTAG);
+ }
if (ibs) {
dn->dn_indblkshift = ibs;
- dn->dn_next_indblkshift[tx->tx_txg&TXG_MASK] = ibs;
+ dn->dn_next_indblkshift[tx->tx_txg & TXG_MASK] = ibs;
}
- /* rele after we have fixed the blocksize in the dnode */
- if (db)
- dbuf_rele(db, FTAG);
rw_exit(&dn->dn_struct_rwlock);
return (0);
return (SET_ERROR(ENOTSUP));
}
+static void
+dnode_set_nlevels_impl(dnode_t *dn, int new_nlevels, dmu_tx_t *tx)
+{
+ uint64_t txgoff = tx->tx_txg & TXG_MASK;
+ int old_nlevels = dn->dn_nlevels;
+ dmu_buf_impl_t *db;
+ list_t *list;
+ dbuf_dirty_record_t *new, *dr, *dr_next;
+
+ ASSERT(RW_WRITE_HELD(&dn->dn_struct_rwlock));
+
+ ASSERT3U(new_nlevels, >, dn->dn_nlevels);
+ dn->dn_nlevels = new_nlevels;
+
+ ASSERT3U(new_nlevels, >, dn->dn_next_nlevels[txgoff]);
+ dn->dn_next_nlevels[txgoff] = new_nlevels;
+
+ /* dirty the left indirects */
+ db = dbuf_hold_level(dn, old_nlevels, 0, FTAG);
+ ASSERT(db != NULL);
+ new = dbuf_dirty(db, tx);
+ dbuf_rele(db, FTAG);
+
+ /* transfer the dirty records to the new indirect */
+ mutex_enter(&dn->dn_mtx);
+ mutex_enter(&new->dt.di.dr_mtx);
+ list = &dn->dn_dirty_records[txgoff];
+ for (dr = list_head(list); dr; dr = dr_next) {
+ dr_next = list_next(&dn->dn_dirty_records[txgoff], dr);
+
+ IMPLY(dr->dr_dbuf == NULL, old_nlevels == 1);
+ if (dr->dr_dbuf == NULL ||
+ (dr->dr_dbuf->db_level == old_nlevels - 1 &&
+ dr->dr_dbuf->db_blkid != DMU_BONUS_BLKID &&
+ dr->dr_dbuf->db_blkid != DMU_SPILL_BLKID)) {
+ list_remove(&dn->dn_dirty_records[txgoff], dr);
+ list_insert_tail(&new->dt.di.dr_children, dr);
+ dr->dr_parent = new;
+ }
+ }
+ mutex_exit(&new->dt.di.dr_mtx);
+ mutex_exit(&dn->dn_mtx);
+}
+
+int
+dnode_set_nlevels(dnode_t *dn, int nlevels, dmu_tx_t *tx)
+{
+ int ret = 0;
+
+ rw_enter(&dn->dn_struct_rwlock, RW_WRITER);
+
+ if (dn->dn_nlevels == nlevels) {
+ ret = 0;
+ goto out;
+ } else if (nlevels < dn->dn_nlevels) {
+ ret = SET_ERROR(EINVAL);
+ goto out;
+ }
+
+ dnode_set_nlevels_impl(dn, nlevels, tx);
+
+out:
+ rw_exit(&dn->dn_struct_rwlock);
+ return (ret);
+}
+
/* read-holding callers must not rely on the lock being continuously held */
void
-dnode_new_blkid(dnode_t *dn, uint64_t blkid, dmu_tx_t *tx, boolean_t have_read)
+dnode_new_blkid(dnode_t *dn, uint64_t blkid, dmu_tx_t *tx, boolean_t have_read,
+ boolean_t force)
{
- uint64_t txgoff = tx->tx_txg & TXG_MASK;
int epbs, new_nlevels;
uint64_t sz;
}
}
- if (blkid <= dn->dn_maxblkid)
+ /*
+ * Raw sends (indicated by the force flag) require that we take the
+ * given blkid even if the value is lower than the current value.
+ */
+ if (!force && blkid <= dn->dn_maxblkid)
goto out;
+ /*
+ * We use the (otherwise unused) top bit of dn_next_maxblkid[txgoff]
+ * to indicate that this field is set. This allows us to set the
+ * maxblkid to 0 on an existing object in dnode_sync().
+ */
dn->dn_maxblkid = blkid;
+ dn->dn_next_maxblkid[tx->tx_txg & TXG_MASK] =
+ blkid | DMU_NEXT_MAXBLKID_SET;
/*
* Compute the number of levels necessary to support the new maxblkid.
+ * Raw sends will ensure nlevels is set correctly for us.
*/
new_nlevels = 1;
epbs = dn->dn_indblkshift - SPA_BLKPTRSHIFT;
sz <= blkid && sz >= dn->dn_nblkptr; sz <<= epbs)
new_nlevels++;
- if (new_nlevels > dn->dn_nlevels) {
- int old_nlevels = dn->dn_nlevels;
- dmu_buf_impl_t *db;
- list_t *list;
- dbuf_dirty_record_t *new, *dr, *dr_next;
+ ASSERT3U(new_nlevels, <=, DN_MAX_LEVELS);
- dn->dn_nlevels = new_nlevels;
-
- ASSERT3U(new_nlevels, >, dn->dn_next_nlevels[txgoff]);
- dn->dn_next_nlevels[txgoff] = new_nlevels;
-
- /* dirty the left indirects */
- db = dbuf_hold_level(dn, old_nlevels, 0, FTAG);
- ASSERT(db != NULL);
- new = dbuf_dirty(db, tx);
- dbuf_rele(db, FTAG);
-
- /* transfer the dirty records to the new indirect */
- mutex_enter(&dn->dn_mtx);
- mutex_enter(&new->dt.di.dr_mtx);
- list = &dn->dn_dirty_records[txgoff];
- for (dr = list_head(list); dr; dr = dr_next) {
- dr_next = list_next(&dn->dn_dirty_records[txgoff], dr);
- if (dr->dr_dbuf->db_level != new_nlevels-1 &&
- dr->dr_dbuf->db_blkid != DMU_BONUS_BLKID &&
- dr->dr_dbuf->db_blkid != DMU_SPILL_BLKID) {
- ASSERT(dr->dr_dbuf->db_level == old_nlevels-1);
- list_remove(&dn->dn_dirty_records[txgoff], dr);
- list_insert_tail(&new->dt.di.dr_children, dr);
- dr->dr_parent = new;
- }
- }
- mutex_exit(&new->dt.di.dr_mtx);
- mutex_exit(&dn->dn_mtx);
+ if (!force) {
+ if (new_nlevels > dn->dn_nlevels)
+ dnode_set_nlevels_impl(dn, new_nlevels, tx);
+ } else {
+ ASSERT3U(dn->dn_nlevels, >=, new_nlevels);
}
out:
}
}
+/*
+ * Dirty all the in-core level-1 dbufs in the range specified by start_blkid
+ * and end_blkid.
+ */
+static void
+dnode_dirty_l1range(dnode_t *dn, uint64_t start_blkid, uint64_t end_blkid,
+ dmu_tx_t *tx)
+{
+ dmu_buf_impl_t *db_search;
+ dmu_buf_impl_t *db;
+ avl_index_t where;
+
+ db_search = kmem_zalloc(sizeof (dmu_buf_impl_t), KM_SLEEP);
+
+ mutex_enter(&dn->dn_dbufs_mtx);
+
+ db_search->db_level = 1;
+ db_search->db_blkid = start_blkid + 1;
+ db_search->db_state = DB_SEARCH;
+ for (;;) {
+
+ db = avl_find(&dn->dn_dbufs, db_search, &where);
+ if (db == NULL)
+ db = avl_nearest(&dn->dn_dbufs, where, AVL_AFTER);
+
+ if (db == NULL || db->db_level != 1 ||
+ db->db_blkid >= end_blkid) {
+ break;
+ }
+
+ /*
+ * Setup the next blkid we want to search for.
+ */
+ db_search->db_blkid = db->db_blkid + 1;
+ ASSERT3U(db->db_blkid, >=, start_blkid);
+
+ /*
+ * If the dbuf transitions to DB_EVICTING while we're trying
+ * to dirty it, then we will be unable to discover it in
+ * the dbuf hash table. This will result in a call to
+ * dbuf_create() which needs to acquire the dn_dbufs_mtx
+ * lock. To avoid a deadlock, we drop the lock before
+ * dirtying the level-1 dbuf.
+ */
+ mutex_exit(&dn->dn_dbufs_mtx);
+ dnode_dirty_l1(dn, db->db_blkid, tx);
+ mutex_enter(&dn->dn_dbufs_mtx);
+ }
+
+#ifdef ZFS_DEBUG
+ /*
+ * Walk all the in-core level-1 dbufs and verify they have been dirtied.
+ */
+ db_search->db_level = 1;
+ db_search->db_blkid = start_blkid + 1;
+ db_search->db_state = DB_SEARCH;
+ db = avl_find(&dn->dn_dbufs, db_search, &where);
+ if (db == NULL)
+ db = avl_nearest(&dn->dn_dbufs, where, AVL_AFTER);
+ for (; db != NULL; db = AVL_NEXT(&dn->dn_dbufs, db)) {
+ if (db->db_level != 1 || db->db_blkid >= end_blkid)
+ break;
+ if (db->db_state != DB_EVICTING)
+ ASSERT(db->db_dirtycnt > 0);
+ }
+#endif
+ kmem_free(db_search, sizeof (dmu_buf_impl_t));
+ mutex_exit(&dn->dn_dbufs_mtx);
+}
+
void
-dnode_free_range(dnode_t *dn, uint64_t off, uint64_t len, dmu_tx_t *tx)
+dnode_set_dirtyctx(dnode_t *dn, dmu_tx_t *tx, const void *tag)
+{
+ /*
+ * Don't set dirtyctx to SYNC if we're just modifying this as we
+ * initialize the objset.
+ */
+ if (dn->dn_dirtyctx == DN_UNDIRTIED) {
+ dsl_dataset_t *ds = dn->dn_objset->os_dsl_dataset;
+
+ if (ds != NULL) {
+ rrw_enter(&ds->ds_bp_rwlock, RW_READER, tag);
+ }
+ if (!BP_IS_HOLE(dn->dn_objset->os_rootbp)) {
+ if (dmu_tx_is_syncing(tx))
+ dn->dn_dirtyctx = DN_DIRTY_SYNC;
+ else
+ dn->dn_dirtyctx = DN_DIRTY_OPEN;
+ dn->dn_dirtyctx_firstset = tag;
+ }
+ if (ds != NULL) {
+ rrw_exit(&ds->ds_bp_rwlock, tag);
+ }
+ }
+}
+
+static void
+dnode_partial_zero(dnode_t *dn, uint64_t off, uint64_t blkoff, uint64_t len,
+ dmu_tx_t *tx)
{
dmu_buf_impl_t *db;
+ int res;
+
+ rw_enter(&dn->dn_struct_rwlock, RW_READER);
+ res = dbuf_hold_impl(dn, 0, dbuf_whichblock(dn, 0, off), TRUE, FALSE,
+ FTAG, &db);
+ rw_exit(&dn->dn_struct_rwlock);
+ if (res == 0) {
+ db_lock_type_t dblt;
+ boolean_t dirty;
+
+ dblt = dmu_buf_lock_parent(db, RW_READER, FTAG);
+ /* don't dirty if not on disk and not dirty */
+ dirty = !list_is_empty(&db->db_dirty_records) ||
+ (db->db_blkptr && !BP_IS_HOLE(db->db_blkptr));
+ dmu_buf_unlock_parent(db, dblt, FTAG);
+ if (dirty) {
+ caddr_t data;
+
+ dmu_buf_will_dirty(&db->db, tx);
+ data = db->db.db_data;
+ memset(data + blkoff, 0, len);
+ }
+ dbuf_rele(db, FTAG);
+ }
+}
+
+void
+dnode_free_range(dnode_t *dn, uint64_t off, uint64_t len, dmu_tx_t *tx)
+{
uint64_t blkoff, blkid, nblks;
int blksz, blkshift, head, tail;
int trunc = FALSE;
int epbs;
- rw_enter(&dn->dn_struct_rwlock, RW_WRITER);
blksz = dn->dn_datablksz;
blkshift = dn->dn_datablkshift;
epbs = dn->dn_indblkshift - SPA_BLKPTRSHIFT;
head = P2NPHASE(off, blksz);
blkoff = P2PHASE(off, blksz);
if ((off >> blkshift) > dn->dn_maxblkid)
- goto out;
+ return;
} else {
ASSERT(dn->dn_maxblkid == 0);
if (off == 0 && len >= blksz) {
/*
* Freeing the whole block; fast-track this request.
- * Note that we won't dirty any indirect blocks,
- * which is fine because we will be freeing the entire
- * file and thus all indirect blocks will be freed
- * by free_children().
*/
blkid = 0;
nblks = 1;
+ if (dn->dn_nlevels > 1) {
+ rw_enter(&dn->dn_struct_rwlock, RW_WRITER);
+ dnode_dirty_l1(dn, 0, tx);
+ rw_exit(&dn->dn_struct_rwlock);
+ }
goto done;
} else if (off >= blksz) {
/* Freeing past end-of-data */
- goto out;
+ return;
} else {
/* Freeing part of the block. */
head = blksz - off;
ASSERT3U(blkoff + head, ==, blksz);
if (len < head)
head = len;
- if (dbuf_hold_impl(dn, 0, dbuf_whichblock(dn, off), TRUE,
- FTAG, &db) == 0) {
- caddr_t data;
-
- /* don't dirty if it isn't on disk and isn't dirty */
- if (db->db_last_dirty ||
- (db->db_blkptr && !BP_IS_HOLE(db->db_blkptr))) {
- rw_exit(&dn->dn_struct_rwlock);
- dmu_buf_will_dirty(&db->db, tx);
- rw_enter(&dn->dn_struct_rwlock, RW_WRITER);
- data = db->db.db_data;
- bzero(data + blkoff, head);
- }
- dbuf_rele(db, FTAG);
- }
+ dnode_partial_zero(dn, off, blkoff, head, tx);
off += head;
len -= head;
}
/* If the range was less than one block, we're done */
if (len == 0)
- goto out;
+ return;
/* If the remaining range is past end of file, we're done */
if ((off >> blkshift) > dn->dn_maxblkid)
- goto out;
+ return;
ASSERT(ISP2(blksz));
if (trunc)
if (tail) {
if (len < tail)
tail = len;
- if (dbuf_hold_impl(dn, 0, dbuf_whichblock(dn, off+len),
- TRUE, FTAG, &db) == 0) {
- /* don't dirty if not on disk and not dirty */
- if (db->db_last_dirty ||
- (db->db_blkptr && !BP_IS_HOLE(db->db_blkptr))) {
- rw_exit(&dn->dn_struct_rwlock);
- dmu_buf_will_dirty(&db->db, tx);
- rw_enter(&dn->dn_struct_rwlock, RW_WRITER);
- bzero(db->db.db_data, tail);
- }
- dbuf_rele(db, FTAG);
- }
+ dnode_partial_zero(dn, off + len, 0, tail, tx);
len -= tail;
}
/* If the range did not include a full block, we are done */
if (len == 0)
- goto out;
+ return;
ASSERT(IS_P2ALIGNED(off, blksz));
ASSERT(trunc || IS_P2ALIGNED(len, blksz));
* amount of space if we copy the freed BPs into deadlists.
*/
if (dn->dn_nlevels > 1) {
- uint64_t first, last, i, ibyte;
- int shift, err;
+ rw_enter(&dn->dn_struct_rwlock, RW_WRITER);
+ uint64_t first, last;
first = blkid >> epbs;
dnode_dirty_l1(dn, first, tx);
if (last != first)
dnode_dirty_l1(dn, last, tx);
- shift = dn->dn_datablkshift + dn->dn_indblkshift -
+ dnode_dirty_l1range(dn, first, last, tx);
+
+ int shift = dn->dn_datablkshift + dn->dn_indblkshift -
SPA_BLKPTRSHIFT;
- for (i = first + 1; i < last; i++) {
+ for (uint64_t i = first + 1; i < last; i++) {
/*
* Set i to the blockid of the next non-hole
* level-1 indirect block at or after i. Note
* that dnode_next_offset() operates in terms of
* level-0-equivalent bytes.
*/
- ibyte = i << shift;
- err = dnode_next_offset(dn, DNODE_FIND_HAVELOCK,
+ uint64_t ibyte = i << shift;
+ int err = dnode_next_offset(dn, DNODE_FIND_HAVELOCK,
&ibyte, 2, 1, 0);
i = ibyte >> shift;
if (i >= last)
dnode_dirty_l1(dn, i, tx);
}
+ rw_exit(&dn->dn_struct_rwlock);
}
done:
*/
mutex_enter(&dn->dn_mtx);
{
- int txgoff = tx->tx_txg & TXG_MASK;
- if (dn->dn_free_ranges[txgoff] == NULL) {
- dn->dn_free_ranges[txgoff] =
- range_tree_create(NULL, NULL, &dn->dn_mtx);
- }
- range_tree_clear(dn->dn_free_ranges[txgoff], blkid, nblks);
- range_tree_add(dn->dn_free_ranges[txgoff], blkid, nblks);
+ int txgoff = tx->tx_txg & TXG_MASK;
+ if (dn->dn_free_ranges[txgoff] == NULL) {
+ dn->dn_free_ranges[txgoff] = range_tree_create(NULL,
+ RANGE_SEG64, NULL, 0, 0);
+ }
+ range_tree_clear(dn->dn_free_ranges[txgoff], blkid, nblks);
+ range_tree_add(dn->dn_free_ranges[txgoff], blkid, nblks);
}
dprintf_dnode(dn, "blkid=%llu nblks=%llu txg=%llu\n",
- blkid, nblks, tx->tx_txg);
+ (u_longlong_t)blkid, (u_longlong_t)nblks,
+ (u_longlong_t)tx->tx_txg);
mutex_exit(&dn->dn_mtx);
dbuf_free_range(dn, blkid, blkid + nblks - 1, tx);
dnode_setdirty(dn, tx);
-out:
-
- rw_exit(&dn->dn_struct_rwlock);
}
static boolean_t
uint64_t
dnode_block_freed(dnode_t *dn, uint64_t blkid)
{
- void *dp = spa_get_dsl(dn->dn_objset->os_spa);
int i;
if (blkid == DMU_BONUS_BLKID)
return (FALSE);
- /*
- * If we're in the process of opening the pool, dp will not be
- * set yet, but there shouldn't be anything dirty.
- */
- if (dp == NULL)
- return (FALSE);
-
if (dn->dn_free_txg)
return (TRUE);
mutex_exit(&dn->dn_mtx);
}
-/*
- * Call when we think we're going to write/free space in open context to track
- * the amount of memory in use by the currently open txg.
- */
-void
-dnode_willuse_space(dnode_t *dn, int64_t space, dmu_tx_t *tx)
-{
- objset_t *os = dn->dn_objset;
- dsl_dataset_t *ds = os->os_dsl_dataset;
- int64_t aspace = spa_get_asize(os->os_spa, space);
-
- if (ds != NULL) {
- dsl_dir_willuse_space(ds->ds_dir, aspace, tx);
- dsl_pool_dirty_space(dmu_tx_pool(tx), space, tx);
- }
-
- dmu_tx_willuse_space(tx, aspace);
-}
-
/*
* Scans a block at the indicated "level" looking for a hole or data,
* depending on 'flags'.
*/
static int
dnode_next_offset_level(dnode_t *dn, int flags, uint64_t *offset,
- int lvl, uint64_t blkfill, uint64_t txg)
+ int lvl, uint64_t blkfill, uint64_t txg)
{
dmu_buf_impl_t *db = NULL;
void *data = NULL;
boolean_t hole;
int i, inc, error, span;
- dprintf("probing object %llu offset %llx level %d of %u\n",
- dn->dn_object, *offset, lvl, dn->dn_phys->dn_nlevels);
+ ASSERT(RW_LOCK_HELD(&dn->dn_struct_rwlock));
hole = ((flags & DNODE_FIND_HOLE) != 0);
inc = (flags & DNODE_FIND_BACKWARDS) ? -1 : 1;
epb = dn->dn_phys->dn_nblkptr;
data = dn->dn_phys->dn_blkptr;
} else {
- uint64_t blkid = dbuf_whichblock(dn, *offset) >> (epbs * lvl);
- error = dbuf_hold_impl(dn, lvl, blkid, TRUE, FTAG, &db);
+ uint64_t blkid = dbuf_whichblock(dn, lvl, *offset);
+ error = dbuf_hold_impl(dn, lvl, blkid, TRUE, FALSE, FTAG, &db);
if (error) {
if (error != ENOENT)
return (error);
*/
return (SET_ERROR(ESRCH));
}
- error = dbuf_read(db, NULL, DB_RF_CANFAIL | DB_RF_HAVESTRUCT);
+ error = dbuf_read(db, NULL,
+ DB_RF_CANFAIL | DB_RF_HAVESTRUCT |
+ DB_RF_NO_DECRYPT | DB_RF_NOPREFETCH);
if (error) {
dbuf_rele(db, FTAG);
return (error);
}
data = db->db.db_data;
+ rw_enter(&db->db_rwlock, RW_READER);
}
-
if (db != NULL && txg != 0 && (db->db_blkptr == NULL ||
- db->db_blkptr->blk_birth <= txg ||
+ BP_GET_LOGICAL_BIRTH(db->db_blkptr) <= txg ||
BP_IS_HOLE(db->db_blkptr))) {
/*
* This can only happen when we are searching up the tree
error = SET_ERROR(ESRCH);
} else if (lvl == 0) {
dnode_phys_t *dnp = data;
- span = DNODE_SHIFT;
+
ASSERT(dn->dn_type == DMU_OT_DNODE);
+ ASSERT(!(flags & DNODE_FIND_BACKWARDS));
- for (i = (*offset >> span) & (blkfill - 1);
- i >= 0 && i < blkfill; i += inc) {
+ for (i = (*offset >> DNODE_SHIFT) & (blkfill - 1);
+ i < blkfill; i += dnp[i].dn_extra_slots + 1) {
if ((dnp[i].dn_type == DMU_OT_NONE) == hole)
break;
- *offset += (1ULL << span) * inc;
}
- if (i < 0 || i == blkfill)
+
+ if (i == blkfill)
error = SET_ERROR(ESRCH);
+
+ *offset = (*offset & ~(DNODE_BLOCK_SIZE - 1)) +
+ (i << DNODE_SHIFT);
} else {
blkptr_t *bp = data;
uint64_t start = *offset;
else
minfill++;
- *offset = *offset >> span;
+ if (span >= 8 * sizeof (*offset)) {
+ /* This only happens on the highest indirection level */
+ ASSERT3U((lvl - 1), ==, dn->dn_phys->dn_nlevels - 1);
+ *offset = 0;
+ } else {
+ *offset = *offset >> span;
+ }
+
for (i = BF64_GET(*offset, 0, epbs);
i >= 0 && i < epb; i += inc) {
if (BP_GET_FILL(&bp[i]) >= minfill &&
BP_GET_FILL(&bp[i]) <= maxfill &&
- (hole || bp[i].blk_birth > txg))
+ (hole || BP_GET_LOGICAL_BIRTH(&bp[i]) > txg))
break;
if (inc > 0 || *offset > 0)
*offset += inc;
}
- *offset = *offset << span;
+
+ if (span >= 8 * sizeof (*offset)) {
+ *offset = start;
+ } else {
+ *offset = *offset << span;
+ }
+
if (inc < 0) {
/* traversing backwards; position offset at the end */
- ASSERT3U(*offset, <=, start);
- *offset = MIN(*offset + (1ULL << span) - 1, start);
+ if (span < 8 * sizeof (*offset))
+ *offset = MIN(*offset + (1ULL << span) - 1,
+ start);
} else if (*offset < start) {
*offset = start;
}
error = SET_ERROR(ESRCH);
}
- if (db)
+ if (db != NULL) {
+ rw_exit(&db->db_rwlock);
dbuf_rele(db, FTAG);
+ }
return (error);
}
return (error);
}
+
+#if defined(_KERNEL)
+EXPORT_SYMBOL(dnode_hold);
+EXPORT_SYMBOL(dnode_rele);
+EXPORT_SYMBOL(dnode_set_nlevels);
+EXPORT_SYMBOL(dnode_set_blksz);
+EXPORT_SYMBOL(dnode_free_range);
+EXPORT_SYMBOL(dnode_evict_dbufs);
+EXPORT_SYMBOL(dnode_evict_bonus);
+#endif
+
+ZFS_MODULE_PARAM(zfs, zfs_, default_bs, INT, ZMOD_RW,
+ "Default dnode block shift");
+ZFS_MODULE_PARAM(zfs, zfs_, default_ibs, INT, ZMOD_RW,
+ "Default dnode indirect block shift");
projects / mirror_zfs.git / blobdiff